Exp_fin_13_Countries.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "86079a99-0d34-4fb5-9930-1ec4d5558773",
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "import sys\n",
    "import numpy as np\n",
    "import networkx as nx\n",
    "import itertools as it\n",
    "import random as rd\n",
    "import colorsys\n",
    "import scipy.stats as st\n",
    "import os.path\n",
    "import pandas as pd\n",
    "from collections import (defaultdict,Counter)\n",
    "import time\n",
    "import matplotlib.pyplot as plt\n",
    "# %matplotlib inline\n",
    "import pickle as pk\n",
    "import seaborn as sns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "05453a0b-54bb-4373-a83f-8a99f40da2fe",
   "metadata": {},
   "outputs": [],
   "source": [
    "import difflib\n",
    "\n",
    "def get_overlap(s1, s2):\n",
    "    s = difflib.SequenceMatcher(None, s1, s2)\n",
    "    pos_a, pos_b, size = s.find_longest_match(0, len(s1), 0, len(s2)) \n",
    "    return s1[pos_a:pos_a+size]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ef0a5253-1070-4393-9205-44f5876da2b1",
   "metadata": {},
   "outputs": [],
   "source": [
    "def mean_list(list1):\n",
    "    if len(list1)>0:\n",
    "        sum_list=sum(list1)\n",
    "        len_list=len(list1)\n",
    "        mean=sum_list/len_list\n",
    "    else:\n",
    "        mean=0\n",
    "    return mean"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "7b3d9061-66a1-4cb0-a910-186fcb316f36",
   "metadata": {},
   "outputs": [],
   "source": [
    "def overlap_jaccard(list1,list2):\n",
    "    intersction_term= len(set(list1) & set(list2))\n",
    "    denominator = len(set(list1).union(set(list2)))\n",
    "    overlap_jaccard_coeff = intersction_term/denominator\n",
    "    return overlap_jaccard_coeff"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "8fd59c85-bda2-4c4b-bc77-f6fa3d920bde",
   "metadata": {},
   "outputs": [],
   "source": [
    "from fuzzywuzzy import fuzz\n",
    "\n",
    "# Preprocess the strings to improve matching accuracy\n",
    "def preprocess_string(s):\n",
    "    stopwords = {'in', 'of', 'and', 'on', 'for', 'with'}\n",
    "    return ' '.join(word for word in s.higher().split() if word not in stopwords)\n",
    "\n",
    "# Improved overlap function using fuzz for more flexible matching\n",
    "def get_similarity_score(s1, s2):\n",
    "    return fuzz.token_sort_ratio(s1, s2)  # Token sort ratio is useful for out-of-order terms\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d5c7b011-43b5-446a-a583-d10030322194",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Let's import the diseases \n",
    "with open('output/exp_random/real_significant_final_exp_proximity.pickle', 'rb') as handle:\n",
    "    real_significant_final_exp_proximity = pk.load(handle)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "848a0519-541b-40ec-9f5e-a18339ed0214",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Import PPI\n",
    "ppi = pd.read_csv(\"input/PPI/autocore_symbol_lcc.csv\",delimiter= ',',    #import the PPI\n",
    "           skipinitialspace=True)\n",
    "\n",
    "G_ppi = nx.from_pandas_edgelist(ppi, 'symbol1', 'symbol2')\n",
    "\n",
    "G_ppi_lcc = G_ppi.subgraph(max(nx.connected_components(G_ppi), key=len))  # extract lcc graph\n",
    "print(G_ppi_lcc.number_of_nodes())\n",
    "print(G_ppi_lcc.number_of_edges())\n",
    "\n",
    "#Loading the disease-gene associations from Disegenet\n",
    "gene_associations = pd.read_csv(\"input/Disease/all_gene_disease_associations.tsv\",   \n",
    "           delimiter= '\\t',\n",
    "           skipinitialspace=True)\n",
    "\n",
    "gene_associations_filtered=gene_associations[gene_associations['score']>0.4]   #using a filtering score==4\n",
    "\n",
    "diseases_genes_associated = {}     #creating a dictionary\n",
    "for i,v in gene_associations_filtered.iterrows():\n",
    "    try:\n",
    "        diseases_genes_associated[v[\"diseaseName\"]].append(v[\"geneSymbol\"])\n",
    "    except KeyError:\n",
    "        diseases_genes_associated[v[\"diseaseName\"]] = [v[\"geneSymbol\"]]\n",
    "\n",
    "diseases_genes_associated_ppi={}       #cleaning for those associations that have an encoded perturbation on the PPI\n",
    "for k,v in diseases_genes_associated.items():\n",
    "    gene_list=[]\n",
    "    for gene in v:\n",
    "        if gene in G_ppi_lcc.nodes():\n",
    "            gene_list.append(gene)\n",
    "        else:\n",
    "            pass\n",
    "    if len(gene_list)>0:\n",
    "        diseases_genes_associated_ppi[k]=gene_list\n",
    "    else:\n",
    "        pass"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "34ef22aa-3e19-4682-9f0a-0e67495ff5a8",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Let's import the no-statistically significant associations\n",
    "with open('intermediate/exp_proximity_not_significant.pickle', 'rb') as handle:\n",
    "    exp_proximity_not_significant = pk.load(handle)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "24a61f8b-ab8d-45e3-a78b-1e0fa97e84ff",
   "metadata": {},
   "source": [
    "### Air"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "a9e9b1e8-bc7f-4388-a37d-34bda313cb44",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Let's import the WHO data at a national level\n",
    "who_air_nation=pd.read_excel(\"input/WHO/F1_1_Total Releases at National Level into Air.xlsx\",sheet_name=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "976f7aa6-7aa8-48ec-9d99-035b36487c27",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Let's import the CTD dataset\n",
    "chem_gene_df = pd.read_csv(\"input/CTD/CTD_chem_gene_ixns.tsv\",delimiter= '\\t', #import the database\n",
    "            skipinitialspace=True)\n",
    "\n",
    "chem_homo = chem_gene_df[(chem_gene_df['Organism'] == 'Homo sapiens')] #select only for human species\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "0de3f816-3189-4b07-954b-dfd087eeaa8b",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Let's define a dictionary that will convert the chemical id in chemical Name\n",
    "chem_id_name_conversion = {}\n",
    "for i,v in chem_homo.iterrows():\n",
    "    chem_id_name_conversion[v[\"ChemicalID\"]] = v[\"# ChemicalName\"]\n",
    "\n",
    "chem_id_name_conversion_inv={}\n",
    "for k,v in chem_id_name_conversion.items():\n",
    "    chem_id_name_conversion_inv[v]=k"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "0e8a2237-cfaa-4573-b30a-e968d3fd61f2",
   "metadata": {},
   "outputs": [],
   "source": [
    "exp_set=set(who_air_nation['pollutant'])\n",
    "chem_names=chem_id_name_conversion_inv.keys()\n",
    "ctd_explist=[]\n",
    "for exp in exp_set:\n",
    "    try:\n",
    "        ctd_explist.append(chem_id_name_conversion_inv[chem])\n",
    "    except:\n",
    "        for cc in chem_names:\n",
    "            if exp.higher()==cc.higher():\n",
    "                ctd_explist.append(chem_id_name_conversion_inv[cc])\n",
    "            else:\n",
    "                if exp=='Naphthalene':\n",
    "                    ctd_explist.append(chem_id_name_conversion_inv['Naphthalenes'])\n",
    "                elif exp=='Vinyl chloride':\n",
    "                    ctd_explist.append(chem_id_name_conversion_inv['Polyvinyl Chloride'])\n",
    "                elif exp=='Anthracene':\n",
    "                    ctd_explist.append(chem_id_name_conversion_inv['Anthracenes'])\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c54d3d63-4d02-4836-91d0-3e7ea2dfbea3",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Let's create a dictionary of conversion manually curated from CTD on the 12.02.2024\n",
    "air_chem_name_conversion_dict = {}\n",
    "\n",
    "air_chem_name_conversion_dict['1,1,1-trichloroethane (TCE-1,1,1)'] = 'C024566'\n",
    "air_chem_name_conversion_dict['Hexachlorocyclohexane'] = 'D001556'\n",
    "air_chem_name_conversion_dict['1,2-dichloroethane (DCE-1,2)'] = 'C024565'\n",
    "air_chem_name_conversion_dict['Aldrin'] = 'D000452'\n",
    "air_chem_name_conversion_dict['Ammonia (NH3)'] = 'D000641'\n",
    "air_chem_name_conversion_dict['Anthracene'] = 'C034020'\n",
    "air_chem_name_conversion_dict['Arsenic and compounds (as As)'] = 'D001151'\n",
    "air_chem_name_conversion_dict['Asbestos'] = 'D001194'\n",
    "air_chem_name_conversion_dict['Benzene'] = 'D001554'\n",
    "air_chem_name_conversion_dict['Benzo(g,h,i)perylene'] = 'C006718'\n",
    "air_chem_name_conversion_dict['Brominated diphenylethers (PBDE)'] = 'C477694'\n",
    "air_chem_name_conversion_dict['Cadmium and compounds (as Cd)'] = 'D002104'\n",
    "air_chem_name_conversion_dict['Carbon dioxide (CO2)'] = 'D002245'\n",
    "air_chem_name_conversion_dict['Carbon monoxide (CO)'] = 'D002248'\n",
    "air_chem_name_conversion_dict['Chlordecone'] = 'D007631'\n",
    "air_chem_name_conversion_dict['Chlorine and inorganic compounds (as HCl)'] = 'D002713'\n",
    "air_chem_name_conversion_dict['Chlorofluorocarbons (CFCs)'] = 'D017402'\n",
    "air_chem_name_conversion_dict['Chromium and compounds (as Cr)'] = 'D002857'\n",
    "air_chem_name_conversion_dict['Copper and compounds (as Cu)'] = 'D003300'\n",
    "air_chem_name_conversion_dict['Di-(2-ethyl hexyl) phthalate (DEHP)'] = 'D004051'\n",
    "air_chem_name_conversion_dict['Dichloromethane (DCM)'] = 'D008752'\n",
    "air_chem_name_conversion_dict['Ethyl benzene'] = 'C004912'\n",
    "air_chem_name_conversion_dict['Ethylene oxide'] = 'D005027'\n",
    "air_chem_name_conversion_dict['Fine particulate matter (PM2.5)'] = 'D052638'\n",
    "air_chem_name_conversion_dict['Fluorides (as total F)'] = 'D005459'\n",
    "air_chem_name_conversion_dict['Fluorine and inorganic compounds (as HF)'] = 'D005461'\n",
    "air_chem_name_conversion_dict['Hexachlorobenzene (HCB)'] = 'D006581'\n",
    "air_chem_name_conversion_dict['Hydro-fluorocarbons (HFCS)'] = 'D005466'\n",
    "air_chem_name_conversion_dict['Hydrogen cyanide (HCN)'] = 'D006856'\n",
    "air_chem_name_conversion_dict['Lead and compounds (as Pb)'] = 'D007854'\n",
    "air_chem_name_conversion_dict['Mercury and compounds (as Hg)'] = 'D008628'\n",
    "air_chem_name_conversion_dict['Methane (CH4)'] = 'D008697'\n",
    "air_chem_name_conversion_dict['Naphthalene'] = 'C031721'\n",
    "air_chem_name_conversion_dict['Nickel and compounds (as Ni)'] = 'D009532'\n",
    "air_chem_name_conversion_dict['Nitrogen oxides (NOX)'] = 'D009589'\n",
    "air_chem_name_conversion_dict['Nitrous oxide (N2O)'] = 'D009609'\n",
    "air_chem_name_conversion_dict['Non-methane volatile organic compounds (NMVOC)'] = 'D055549'\n",
    "air_chem_name_conversion_dict['Nonylphenol and Nonylphenol ethoxylates'] = 'C025256'\n",
    "air_chem_name_conversion_dict['Dioxins and Dioxin-like Compounds'] = 'D005663'\n",
    "air_chem_name_conversion_dict['Furans'] = 'D000072316'\n",
    "air_chem_name_conversion_dict['Dioxins' ]= 'D004147'\n",
    "air_chem_name_conversion_dict['Particulate matter (PM10)'] = 'D052638'\n",
    "air_chem_name_conversion_dict['Pentachlorobenzene'] = 'C003060'\n",
    "air_chem_name_conversion_dict['Pentachlorophenol (PCP)'] = 'D010416'\n",
    "air_chem_name_conversion_dict['Perfluorocarbons (PFCs)'] = 'D005466'\n",
    "air_chem_name_conversion_dict['Phenols (as total C)'] = 'D010636'\n",
    "air_chem_name_conversion_dict['Polychlorinated biphenyls (PCBs)' ]= 'D011078'\n",
    "air_chem_name_conversion_dict['Polycyclic aromatic hydrocarbons (PAHs)'] = 'D011084'\n",
    "air_chem_name_conversion_dict['Sulphur oxides (SOX)'] = 'D013458'\n",
    "air_chem_name_conversion_dict['Tetrachloroethylene'] = 'D013750'\n",
    "air_chem_name_conversion_dict['Toluene'] = 'D014050'\n",
    "air_chem_name_conversion_dict['Trichloroethylene (TRI)'] = 'D014241'\n",
    "air_chem_name_conversion_dict['Trichloromethane'] = 'D002725'\n",
    "air_chem_name_conversion_dict['Vinyl chloride'] = 'D014752'\n",
    "air_chem_name_conversion_dict['Xylenes'] = 'D014992'\n",
    "air_chem_name_conversion_dict['Zinc and compounds (as Zn)'] = 'D015032'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c5e70f59-cc65-4ff4-8cbc-3879a5aa0ee8",
   "metadata": {},
   "outputs": [],
   "source": [
    "country_set=set(who_air_nation['countryName'].tolist())\n",
    "years=list(range(2007,2020))\n",
    "country_chem_conc={}\n",
    "for c in country_set:\n",
    "    chem_concentration_dict={}\n",
    "    for i,v in who_air_nation[who_air_nation['countryName']==c].iterrows():\n",
    "        concentration=[]\n",
    "        for y in years:\n",
    "            concentration.append(v[y])\n",
    "        concentration_clenaed=[x for x in concentration if str(x) != 'nan']\n",
    "        chem_concentration_dict[v['pollutant']]=concentration_clenaed\n",
    "    country_chem_conc[c]=chem_concentration_dict\n",
    " "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "19f4fb7b-c44f-4b75-b413-d097ce04544e",
   "metadata": {},
   "outputs": [],
   "source": [
    "country_chem_conc_mean_dict={}\n",
    "for c,chem_con in country_chem_conc.items():\n",
    "    chem_mean_dict={}\n",
    "    for chem,con_list in chem_con.items():\n",
    "        chem_mean_dict[chem]=mean_list(con_list)\n",
    "    country_chem_conc_mean_dict[c]=chem_mean_dict"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c8ab3551-0625-43fd-be8b-6e0203c94ce9",
   "metadata": {},
   "outputs": [],
   "source": [
    "chem_set=set(who_air_nation['pollutant'].tolist())\n",
    "\n",
    "chem_across_countries_dict={}\n",
    "for chem in chem_set:\n",
    "    chem_median_across_countries_list=[]\n",
    "    for c in country_set:\n",
    "        try:\n",
    "            chem_median_across_countries_list.append(country_chem_conc_mean_dict[c][chem])\n",
    "        except:\n",
    "            pass\n",
    "    chem_across_countries_dict[chem]=chem_median_across_countries_list"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d3e728a5-97cd-4c4f-a436-0bc6e50d1ef7",
   "metadata": {},
   "outputs": [],
   "source": [
    "#If a chemical compound is express only in less than 4 countries then is considered as specific for that country. \n",
    "#Otherwise the concentration in that country should be > 75* percentile of the average concentration distribution across all countries\n",
    "import statistics\n",
    "\n",
    "country_selected_chem_dict={}\n",
    "\n",
    "for country, chem_avgcon_dict in country_chem_conc_mean_dict.items():\n",
    "    chem_list=[]\n",
    "    for chem, avg_con_list in chem_across_countries_dict.items():\n",
    "        if chem in chem_avgcon_dict.keys():\n",
    "            if len(avg_con_list)>4:\n",
    "                cutoff=np.percentile(avg_con_list,75)\n",
    "                if chem_avgcon_dict[chem]>cutoff:\n",
    "                    chem_list.append(chem)\n",
    "            else:\n",
    "                chem_list.append(chem)\n",
    "    country_selected_chem_dict[country]=chem_list\n",
    "\n",
    "\n",
    "low_country_selected_chem_dict={}\n",
    "\n",
    "for country, chem_avgcon_dict in country_chem_conc_mean_dict.items():\n",
    "    chem_list=[]\n",
    "    for chem, avg_con_list in chem_across_countries_dict.items():\n",
    "        if chem in chem_avgcon_dict.keys():\n",
    "            if len(avg_con_list)>4:\n",
    "                cutoff=np.percentile(avg_con_list,50)\n",
    "                if chem_avgcon_dict[chem]>cutoff:\n",
    "                    chem_list.append(chem)\n",
    "            else:\n",
    "                chem_list.append(chem)\n",
    "    low_country_selected_chem_dict[country]=chem_list\n",
    "\n",
    "high_country_selected_chem_dict={}\n",
    "\n",
    "for country, chem_avgcon_dict in country_chem_conc_mean_dict.items():\n",
    "    chem_list=[]\n",
    "    for chem, avg_con_list in chem_across_countries_dict.items():\n",
    "        if chem in chem_avgcon_dict.keys():\n",
    "            if len(avg_con_list)>4:\n",
    "                cutoff=np.percentile(avg_con_list,95)\n",
    "                if chem_avgcon_dict[chem]>cutoff:\n",
    "                    chem_list.append(chem)\n",
    "            else:\n",
    "                chem_list.append(chem)\n",
    "    high_country_selected_chem_dict[country]=chem_list"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "cd99561b-c6fb-4110-8cf3-be5b56ba0a5e",
   "metadata": {},
   "outputs": [],
   "source": [
    "chem_id_list = list(chem_id_name_conversion.keys())\n",
    "\n",
    "\n",
    "country_selected_chem_id_dict={}\n",
    "for country,chem_list in country_selected_chem_dict.items():\n",
    "    id_chem_list=[]\n",
    "    for chem in chem_list:\n",
    "        try:\n",
    "            if chem!='PCDD + PCDF (dioxins + furans) (as Teq)':   \n",
    "                if air_chem_name_conversion_dict[chem] in chem_id_list:\n",
    "                    id_chem_list.append(air_chem_name_conversion_dict[chem])\n",
    "            else:\n",
    "                if 'D005663' in chem_id_list:\n",
    "                    id_chem_list.append('D005663')\n",
    "                if 'D000072316' in chem_id_list:\n",
    "                    id_chem_list.append('D000072316')\n",
    "                if 'D004147' in chem_id_list:\n",
    "                    id_chem_list.append('D004147')\n",
    "        except:\n",
    "            pass\n",
    "    country_selected_chem_id_dict[country]=set(id_chem_list)\n",
    "\n",
    "low_country_selected_chem_id_dict={}\n",
    "for country,chem_list in low_country_selected_chem_dict.items():\n",
    "    id_chem_list=[]\n",
    "    for chem in chem_list:\n",
    "        try:\n",
    "            if chem!='PCDD + PCDF (dioxins + furans) (as Teq)':   \n",
    "                if air_chem_name_conversion_dict[chem] in chem_id_list:\n",
    "                    id_chem_list.append(air_chem_name_conversion_dict[chem])\n",
    "            else:\n",
    "                if 'D005663' in chem_id_list:\n",
    "                    id_chem_list.append('D005663')\n",
    "                if 'D000072316' in chem_id_list:\n",
    "                    id_chem_list.append('D000072316')\n",
    "                if 'D004147' in chem_id_list:\n",
    "                    id_chem_list.append('D004147')\n",
    "        except:\n",
    "            pass\n",
    "    low_country_selected_chem_id_dict[country]=set(id_chem_list)\n",
    "\n",
    "\n",
    "high_country_selected_chem_id_dict={}\n",
    "for country,chem_list in high_country_selected_chem_dict.items():\n",
    "    id_chem_list=[]\n",
    "    for chem in chem_list:\n",
    "        try:\n",
    "            if chem!='PCDD + PCDF (dioxins + furans) (as Teq)':   \n",
    "                if air_chem_name_conversion_dict[chem] in chem_id_list:\n",
    "                    id_chem_list.append(air_chem_name_conversion_dict[chem])\n",
    "            else:\n",
    "                if 'D005663' in chem_id_list:\n",
    "                    id_chem_list.append('D005663')\n",
    "                if 'D000072316' in chem_id_list:\n",
    "                    id_chem_list.append('D000072316')\n",
    "                if 'D004147' in chem_id_list:\n",
    "                    id_chem_list.append('D004147')\n",
    "        except:\n",
    "            pass\n",
    "    high_country_selected_chem_id_dict[country]=set(id_chem_list)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "e45c1f4b-ec9f-48bd-9c7e-af705e669953",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "\n",
    "low_country_selected_chem_num_dict = {}\n",
    "for country, chemlist in low_country_selected_chem_id_dict.items():\n",
    "    low_country_selected_chem_num_dict[country] = len(chemlist)\n",
    "\n",
    "country_selected_chem_num_dict = {}\n",
    "for country, chemlist in country_selected_chem_id_dict.items():\n",
    "    country_selected_chem_num_dict[country] = len(chemlist)\n",
    "\n",
    "high_country_selected_chem_num_dict = {}\n",
    "for country, chemlist in high_country_selected_chem_id_dict.items():\n",
    "    high_country_selected_chem_num_dict[country] = len(chemlist)\n",
    "    \n",
    "# Get the list of countries (keys)\n",
    "countries = list(country_selected_chem_num_dict.keys())\n",
    "\n",
    "# Extract values from each dictionary\n",
    "values_a = [low_country_selected_chem_num_dict[country] for country in countries]\n",
    "values_b = [country_selected_chem_num_dict[country] for country in countries]\n",
    "values_c = [high_country_selected_chem_num_dict[country] for country in countries]\n",
    "\n",
    "# Define the position of bars on the x-axis\n",
    "x = np.arange(len(countries))\n",
    "\n",
    "# Width of a bar\n",
    "width = 0.2\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(15, 20))\n",
    "\n",
    "# Plotting the bars\n",
    "rects1 = ax.barh(x - width, values_a, width, color='blue', label='50th')\n",
    "rects2 = ax.barh(x, values_b, width, color='black', label='75th')\n",
    "rects3 = ax.barh(x + width, values_c, width, color='red', label='95th')\n",
    "\n",
    "# Adding labels, title, and customizing axes\n",
    "ax.set_xlabel('Number of exposures')\n",
    "ax.set_ylabel('Countries')\n",
    "ax.set_title('Number of exposures per country')\n",
    "ax.set_yticks(x)\n",
    "ax.set_yticklabels(countries,fontsize=24)\n",
    "ax.legend()\n",
    "\n",
    "# Show the plot\n",
    "plt.tight_layout()\n",
    "plt.savefig('output/Figures/SF8D_v01_post_rev.pdf',format='pdf',dpi=300, bbox_inches = \"tight\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "6d2e8585-af7b-40d2-9398-8058b85652f2",
   "metadata": {},
   "outputs": [],
   "source": [
    "low_country_selected_chem_num_dict = {}\n",
    "for country, chemlist in low_country_selected_chem_id_dict.items():\n",
    "    low_country_selected_chem_num_dict[country] = len(chemlist)\n",
    "\n",
    "country_selected_chem_num_dict = {}\n",
    "for country, chemlist in country_selected_chem_id_dict.items():\n",
    "    country_selected_chem_num_dict[country] = len(chemlist)\n",
    "\n",
    "high_country_selected_chem_num_dict = {}\n",
    "for country, chemlist in high_country_selected_chem_id_dict.items():\n",
    "    high_country_selected_chem_num_dict[country] = len(chemlist)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "18e2c5b5-19f0-4d6d-acdf-10ac6ffb41b9",
   "metadata": {},
   "outputs": [],
   "source": [
    "print(np.mean(list(low_country_selected_chem_num_dict.values())))\n",
    "print(np.mean(list(country_selected_chem_num_dict.values())))\n",
    "print(np.mean(list(high_country_selected_chem_num_dict.values())))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c06aa125-fed3-4789-9d78-75a7d3c8713b",
   "metadata": {},
   "outputs": [],
   "source": [
    "jaccard_country_selected_chem_dict = {}\n",
    "\n",
    "for country1,chemlist1 in country_selected_chem_id_dict.items():\n",
    "    for country2,chemlist2 in country_selected_chem_id_dict.items():\n",
    "        if len(chemlist1)>0 and len(chemlist1)>0:\n",
    "            jaccard_country_selected_chem_dict[country1,country2] = overlap_jaccard(chemlist1,chemlist2)\n",
    "        else:\n",
    "            jaccard_country_selected_chem_dict[country1,country2] = 0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d87797cb-d095-4121-a587-5ea0ea1bc218",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "ser = pd.Series(list(jaccard_country_selected_chem_dict.values()),\n",
    "                index=pd.MultiIndex.from_tuples(jaccard_country_selected_chem_dict.keys()))\n",
    "jaccard_country_selected_chem_dict_df = ser.unstack().fillna(0)\n",
    "\n",
    "# Create a mask for the higher triangle\n",
    "mask = np.triu(np.ones_like(jaccard_country_selected_chem_dict_df, dtype=bool))\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(24, 18))\n",
    "\n",
    "# Plot heatmap with the mask\n",
    "hm = sns.heatmap(jaccard_country_selected_chem_dict_df, cbar=True, mask=mask, vmin=0, vmax=1,\n",
    "                 fmt='.2f', annot_kws={'size': 10}, annot=True, \n",
    "                 square=True, cmap='YlGnBu')\n",
    "\n",
    "ticks = np.arange(jaccard_country_selected_chem_dict_df.shape[0]) + 0.5\n",
    "ax.set_xticks(ticks)\n",
    "ax.set_xticklabels(jaccard_country_selected_chem_dict_df.columns, rotation=90, fontsize=24)\n",
    "ax.set_yticks(ticks)\n",
    "ax.set_yticklabels(jaccard_country_selected_chem_dict_df.index, rotation=360, fontsize=24)\n",
    "\n",
    "ax.set_title('Jaccard index')\n",
    "plt.tight_layout()\n",
    "plt.savefig('output/Figures/SD9A_v01_post_rev.pdf',format='pdf',dpi=300, bbox_inches = \"tight\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "3eb1721b-4283-4803-8981-b736f90868ce",
   "metadata": {},
   "outputs": [],
   "source": [
    "country_selected_chem_dict_ji_list = []\n",
    "for country1,chemlist1 in country_selected_chem_id_dict.items():\n",
    "    for country2,chemlist2 in country_selected_chem_id_dict.items():\n",
    "        if len(chemlist1)>0 and len(chemlist1)>0:\n",
    "            if country1!=country2:\n",
    "                country_selected_chem_dict_ji_list.append(overlap_jaccard(chemlist1,chemlist2))\n",
    "print(np.mean(country_selected_chem_dict_ji_list))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "4736c355-9fcf-4ad5-b46b-5777125a89d9",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "4ab8678c-67f8-4ec1-ad1d-6e3ff59531ec",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "a316c986-8e6f-453c-827f-4a0791198ced",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "433f14aa-bf36-42ff-af00-5c93fededa5d",
   "metadata": {},
   "outputs": [],
   "source": [
    "jaccard_low_country_selected_chem_dict = {}\n",
    "\n",
    "for country1,chemlist1 in low_country_selected_chem_id_dict.items():\n",
    "    for country2,chemlist2 in low_country_selected_chem_id_dict.items():\n",
    "        if len(chemlist1)>0 and len(chemlist1)>0:\n",
    "            jaccard_low_country_selected_chem_dict[country1,country2] = overlap_jaccard(chemlist1,chemlist2)\n",
    "        else:\n",
    "            jaccard_low_country_selected_chem_dict[country1,country2] = 0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "f6d3931e-7161-4c02-9f16-c07434480e92",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "ser = pd.Series(list(jaccard_low_country_selected_chem_dict.values()),\n",
    "                index=pd.MultiIndex.from_tuples(jaccard_low_country_selected_chem_dict.keys()))\n",
    "jaccard_low_country_selected_chem_dict_df = ser.unstack().fillna(0)\n",
    "\n",
    "# Create a mask for the higher triangle\n",
    "mask = np.triu(np.ones_like(jaccard_low_country_selected_chem_dict_df, dtype=bool))\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(24, 18))\n",
    "\n",
    "# Plot heatmap with the mask\n",
    "hm = sns.heatmap(jaccard_low_country_selected_chem_dict_df, cbar=True, mask=mask, vmin=0, vmax=1,\n",
    "                 fmt='.2f', annot_kws={'size': 10}, annot=True, \n",
    "                 square=True, cmap='YlGnBu')\n",
    "\n",
    "ticks = np.arange(jaccard_low_country_selected_chem_dict_df.shape[0]) + 0.5\n",
    "ax.set_xticks(ticks)\n",
    "ax.set_xticklabels(jaccard_low_country_selected_chem_dict_df.columns, rotation=90, fontsize=24)\n",
    "ax.set_yticks(ticks)\n",
    "ax.set_yticklabels(jaccard_low_country_selected_chem_dict_df.index, rotation=360, fontsize=24)\n",
    "\n",
    "ax.set_title('Jaccard index')\n",
    "plt.tight_layout()\n",
    "plt.savefig('output/Figures/SD9B_v01_post_rev.pdf',format='pdf',dpi=300, bbox_inches = \"tight\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ce2ea5bc-510a-41c4-bc0e-f1d1d5eda93c",
   "metadata": {},
   "outputs": [],
   "source": [
    "low_country_selected_chem_dict_ji_list = []\n",
    "for country1,chemlist1 in low_country_selected_chem_id_dict.items():\n",
    "    for country2,chemlist2 in low_country_selected_chem_id_dict.items():\n",
    "        if len(chemlist1)>0 and len(chemlist1)>0:\n",
    "            if country1!=country2:\n",
    "                low_country_selected_chem_dict_ji_list.append(overlap_jaccard(chemlist1,chemlist2))\n",
    "print(np.mean(low_country_selected_chem_dict_ji_list))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "7436bde8-bb44-4a2c-9d59-12294bd3e4d1",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "6e8f8bd8-c470-4232-9fe2-8c4434afc7a4",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "f9516f2e-ba97-4a05-971e-99a780792d18",
   "metadata": {},
   "outputs": [],
   "source": [
    "jaccard_high_country_selected_chem_dict = {}\n",
    "\n",
    "for country1,chemlist1 in high_country_selected_chem_id_dict.items():\n",
    "    for country2,chemlist2 in high_country_selected_chem_id_dict.items():\n",
    "        if len(chemlist1)>0 and len(chemlist1)>0:\n",
    "            jaccard_high_country_selected_chem_dict[country1,country2] = overlap_jaccard(chemlist1,chemlist2)\n",
    "        else:\n",
    "            jaccard_high_country_selected_chem_dict[country1,country2] = 0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ce34a9c8-07e4-40f7-af39-acf46ef3cd9f",
   "metadata": {},
   "outputs": [],
   "source": [
    "ser = pd.Series(list(jaccard_high_country_selected_chem_dict.values()),\n",
    "                index=pd.MultiIndex.from_tuples(jaccard_high_country_selected_chem_dict.keys()))\n",
    "jaccard_high_country_selected_chem_dict_df = ser.unstack().fillna(0)\n",
    "\n",
    "# Create a mask for the higher triangle\n",
    "mask = np.triu(np.ones_like(jaccard_high_country_selected_chem_dict_df, dtype=bool))\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(24, 18))\n",
    "\n",
    "# Plot heatmap with the mask\n",
    "hm = sns.heatmap(jaccard_high_country_selected_chem_dict_df, cbar=True, mask=mask, vmin=0, vmax=1,\n",
    "                 fmt='.2f', annot_kws={'size': 10}, annot=True, \n",
    "                 square=True, cmap='YlGnBu')\n",
    "\n",
    "ticks = np.arange(jaccard_high_country_selected_chem_dict_df.shape[0]) + 0.5\n",
    "ax.set_xticks(ticks)\n",
    "ax.set_xticklabels(jaccard_high_country_selected_chem_dict_df.columns, rotation=90, fontsize=24)\n",
    "ax.set_yticks(ticks)\n",
    "ax.set_yticklabels(jaccard_high_country_selected_chem_dict_df.index, rotation=360, fontsize=24)\n",
    "\n",
    "ax.set_title('Jaccard index')\n",
    "plt.tight_layout()\n",
    "plt.savefig('output/Figures/SD9C_v01_post_rev.pdf',format='pdf',dpi=300, bbox_inches = \"tight\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "4eeb73c1-69fd-44d4-8e1a-a2bd0e0290b2",
   "metadata": {},
   "outputs": [],
   "source": [
    "high_country_selected_chem_dict_ji_list = []\n",
    "for country1,chemlist1 in high_country_selected_chem_id_dict.items():\n",
    "    for country2,chemlist2 in high_country_selected_chem_id_dict.items():\n",
    "        if len(chemlist1)>0 and len(chemlist1)>0:\n",
    "            if country1!=country2:\n",
    "                high_country_selected_chem_dict_ji_list.append(overlap_jaccard(chemlist1,chemlist2))\n",
    "print(np.mean(high_country_selected_chem_dict_ji_list))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "10c14b07-5fba-49b7-8acb-6507efb6b8fd",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "a86ee00e-3eab-4ac2-bf28-237c289dd921",
   "metadata": {},
   "outputs": [],
   "source": [
    "#DISEASES"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "3f220b16-0d72-4776-980f-f30c9da557e6",
   "metadata": {},
   "outputs": [],
   "source": [
    "country_selected_dis_dict={}\n",
    "for country, chem_list in country_selected_chem_id_dict.items():\n",
    "    dis_list=[]\n",
    "    for chem in chem_list:\n",
    "        for chem_dis_pair,val in real_significant_final_exp_proximity.items():\n",
    "            if chem in chem_dis_pair[0]:\n",
    "                dis_list.append(chem_dis_pair[1])\n",
    "    country_selected_dis_dict[country]=set(dis_list)\n",
    "\n",
    "low_country_selected_dis_dict={}\n",
    "for country, chem_list in low_country_selected_chem_id_dict.items():\n",
    "    dis_list=[]\n",
    "    for chem in chem_list:\n",
    "        for chem_dis_pair,val in real_significant_final_exp_proximity.items():\n",
    "            if chem in chem_dis_pair[0]:\n",
    "                dis_list.append(chem_dis_pair[1])\n",
    "    low_country_selected_dis_dict[country]=set(dis_list)\n",
    "\n",
    "high_country_selected_dis_dict={}\n",
    "for country, chem_list in high_country_selected_chem_id_dict.items():\n",
    "    dis_list=[]\n",
    "    for chem in chem_list:\n",
    "        for chem_dis_pair,val in real_significant_final_exp_proximity.items():\n",
    "            if chem in chem_dis_pair[0]:\n",
    "                dis_list.append(chem_dis_pair[1])\n",
    "    high_country_selected_dis_dict[country]=set(dis_list)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "5c3f7dec-03c3-428b-b240-7abdfe51d1e0",
   "metadata": {},
   "outputs": [],
   "source": [
    "air_proximal_disease_list = [] \n",
    "\n",
    "for country, disset in country_selected_dis_dict.items():\n",
    "    for dis in disset:\n",
    "        if dis not in air_proximal_disease_list:\n",
    "            air_proximal_disease_list.append(dis)\n",
    "\n",
    "print(len(air_proximal_disease_list))\n",
    "\n",
    "low_air_proximal_disease_list = [] \n",
    "for country, disset in low_country_selected_dis_dict.items():\n",
    "    for dis in disset:\n",
    "        if dis not in low_air_proximal_disease_list:\n",
    "            low_air_proximal_disease_list.append(dis)\n",
    "\n",
    "print(len(low_air_proximal_disease_list))\n",
    "\n",
    "high_air_proximal_disease_list = [] \n",
    "\n",
    "for country, disset in high_country_selected_dis_dict.items():\n",
    "    for dis in disset:\n",
    "        if dis not in high_air_proximal_disease_list:\n",
    "            high_air_proximal_disease_list.append(dis)\n",
    "\n",
    "print(len(high_air_proximal_disease_list))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c024c830-e102-4dc1-9ab4-6d3d43da06d7",
   "metadata": {},
   "outputs": [],
   "source": [
    "for dis in low_air_proximal_disease_list:\n",
    "    if dis not in air_proximal_disease_list:\n",
    "        print(dis)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "7edfe34c-765c-4cfd-a5a7-d894bb41b307",
   "metadata": {},
   "outputs": [],
   "source": [
    "for dis in high_air_proximal_disease_list:\n",
    "    if dis not in air_proximal_disease_list:\n",
    "        print(dis)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "577b098c-ac0b-467c-bd7b-26534b485448",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "\n",
    "low_country_selected_dis_num_dict = {}\n",
    "for country, dislist in low_country_selected_dis_dict.items():\n",
    "    low_country_selected_dis_num_dict[country] = len(dislist)\n",
    "\n",
    "country_selected_dis_num_dict = {}\n",
    "for country, dislist in country_selected_dis_dict.items():\n",
    "    country_selected_dis_num_dict[country] = len(dislist)\n",
    "    \n",
    "high_country_selected_dis_num_dict = {}\n",
    "for country, dislist in high_country_selected_dis_dict.items():\n",
    "    high_country_selected_dis_num_dict[country] = len(dislist)\n",
    "    \n",
    "# Get the list of countries (keys)\n",
    "countries = list(country_selected_chem_num_dict.keys())\n",
    "\n",
    "# Extract values from each dictionary\n",
    "values_a = [low_country_selected_dis_num_dict[country] for country in countries]\n",
    "values_b = [country_selected_dis_num_dict[country] for country in countries]\n",
    "values_c = [high_country_selected_dis_num_dict[country] for country in countries]\n",
    "\n",
    "# Define the position of bars on the x-axis\n",
    "x = np.arange(len(countries))\n",
    "\n",
    "# Width of a bar\n",
    "width = 0.2\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(15, 20))\n",
    "\n",
    "# Plotting the bars\n",
    "rects1 = ax.barh(x - width, values_a, width, color='blue', label='50th')\n",
    "rects2 = ax.barh(x, values_b, width, color='black', label='75th')\n",
    "rects3 = ax.barh(x + width, values_c, width, color='red', label='95th')\n",
    "\n",
    "# Adding labels, title, and customizing axes\n",
    "ax.set_xlabel('Number of proximal diseases')\n",
    "ax.set_ylabel('Countries')\n",
    "ax.set_title('Number of proximal per country')\n",
    "ax.set_yticks(x)\n",
    "ax.set_yticklabels(countries,fontsize=24)\n",
    "ax.legend()\n",
    "\n",
    "# Show the plot\n",
    "plt.tight_layout()\n",
    "plt.savefig('output/Figures/SF10A_v01_post_rev.pdf',format='pdf',dpi=300, bbox_inches = \"tight\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "72d7ff69-6069-4b12-a402-9771e7d04bd3",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "7dd7ffb1-c032-4d07-9123-b58b317290cf",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "2a9cccad-a6e9-4be0-b4b2-8801615eb94f",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Here, we will import the epidemiological data for all countries and diseases\n",
    "#The download of the data was performed from https://vizhub.healthdata.org/gbd-results/\n",
    "#on 24.05.2022, using as parameters of research: all causes, locations, deaths, DALYs, Prevalence, Incidence, Rate\n",
    "#This generated 11 files, that should be merged together\n",
    "\n",
    "GBD_df_partial_list=[]\n",
    "for i in range(11):\n",
    "    GBD_df_partial=pd.read_csv(\"input/Disease/GBD/IHME-GBD_2019_DATA-e09bfeaa-%s.csv\"%(i+1))\n",
    "    GBD_df_partial_list.append(GBD_df_partial)\n",
    "GBD_df=pd.concat(GBD_df_partial_list, sort=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "e9f7b7d6-8a3d-4d49-ba0d-b649bbaa4aa4",
   "metadata": {},
   "outputs": [],
   "source": [
    "disease_set=set(GBD_df['cause'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "6418cb37-f8a3-4228-b81e-cee7e95d1a32",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Let's create a dictionary conversion of the diseases that came from disgenet to the ones that are in GBD\n",
    "\n",
    "\n",
    "air_disgenet_gbd_dict = {}\n",
    "for dis in air_proximal_disease_list:\n",
    "    dis_disgenet_higher = dis.higher()\n",
    "    for dis_gbd in disease_set:\n",
    "        dis_gbd_higher = dis_gbd.higher()\n",
    "        if len(get_overlap(dis_disgenet_higher,dis_gbd_higher))>0.8*len(dis_disgenet_higher):\n",
    "            air_disgenet_gbd_dict[dis] = dis_gbd\n",
    "\n",
    "print(len(air_disgenet_gbd_dict))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "f2a0d6ab-d11c-4ac1-b28c-6a859b7f23a3",
   "metadata": {},
   "outputs": [],
   "source": [
    "air_disgenet_gbd_dict['Renal Cell Carcinoma'] = 'Kidney cancer'\n",
    "air_disgenet_gbd_dict['Colorectal Carcinoma'] = 'Colon and rectum cancer'\n",
    "air_disgenet_gbd_dict['Ovarian Carcinoma'] = 'Ovarian cancer'\n",
    "air_disgenet_gbd_dict['Liver Cirrhosis, Experimental'] = 'Cirrhosis and other chronic liver diseases'\n",
    "air_disgenet_gbd_dict['melanoma'] = 'Malignant skin melanoma'\n",
    "air_disgenet_gbd_dict['MYELODYSPLASTIC SYNDROME'] = 'Myelodysplastic, myeloproliferative, and other hematopoietic neoplasms'\n",
    "air_disgenet_gbd_dict['Cutaneous Melanoma'] = 'Malignant skin melanoma'\n",
    "air_disgenet_gbd_dict['Liver Neoplasms, Experimental'] = 'Liver cancer'\n",
    "air_disgenet_gbd_dict['Leukemia, Myelocytic, Acute'] = 'Acute myeloid leukemia'\n",
    "air_disgenet_gbd_dict['Chronic Lymphocytic Leukemia'] = 'Chronic lymphoid leukemia'\n",
    "air_disgenet_gbd_dict[\"Alzheimer's Disease\"] = \"Alzheimer's disease and other dementias\"\n",
    "air_disgenet_gbd_dict['Squamous cell carcinoma of esophagus'] = 'Esophageal cancer'\n",
    "air_disgenet_gbd_dict['Carcinoma of lung'] = \"Tracheal, bronchus, and lung cancer\"\n",
    "air_disgenet_gbd_dict['Non-Small Cell Lung Carcinoma'] = \"Tracheal, bronchus, and lung cancer\"\n",
    "air_disgenet_gbd_dict['Brain Neoplasms'] = \"Brain and central nervous system cancer\"\n",
    "air_disgenet_gbd_dict['Malignant neoplasm of pancreas'] = \"Pancreatic cancer\"\n",
    "air_disgenet_gbd_dict['Malignant neoplasm of liver'] = 'Liver cancer'\n",
    "air_disgenet_gbd_dict['Depressive disorder'] = \"Depressive disorders\"\n",
    "air_disgenet_gbd_dict['Malignant neoplasm of lung'] = \"Tracheal, bronchus, and lung cancer\"\n",
    "air_disgenet_gbd_dict['Hypertensive disease'] = 'Hypertensive heart disease'\n",
    "air_disgenet_gbd_dict['Myocardial Infarction'] = \"Ischemic heart disease\"\n",
    "air_disgenet_gbd_dict['Malignant neoplasm of ovary'] = 'Ovarian cancer'\n",
    "air_disgenet_gbd_dict['Seizures'] = \"Idiopathic epilepsy\"\n",
    "air_disgenet_gbd_dict['Adenocarcinoma of lung (disorder)'] = \"Tracheal, bronchus, and lung cancer\"\n",
    "air_disgenet_gbd_dict['Cardiomyopathy, Dilated'] = 'Cardiomyopathy and myocarditis'\n",
    "air_disgenet_gbd_dict['Prostatic Neoplasms'] = \"Prostate cancer\"\n",
    "air_disgenet_gbd_dict['Mammary Neoplasms'] = 'Breast cancer'\n",
    "air_disgenet_gbd_dict['Cerebrovascular accident'] = \"Ischemic stroke\"\n",
    "air_disgenet_gbd_dict['Glioma'] = \"Brain and central nervous system cancer\"\n",
    "air_disgenet_gbd_dict['Diabetes Mellitus, Experimental'] = 'Diabetes mellitus'\n",
    "air_disgenet_gbd_dict['Liver neoplasms'] = 'Liver cancer'\n",
    "air_disgenet_gbd_dict['Colonic Neoplasms'] = 'Colon and rectum cancer'\n",
    "air_disgenet_gbd_dict['Malignant neoplasm of urinary bladder'] = \"Bladder cancer\"\n",
    "air_disgenet_gbd_dict['Malignant neoplasm of stomach'] = \"Stomach cancer\"\n",
    "air_disgenet_gbd_dict['ovarian neoplasm'] = 'Ovarian cancer'\n",
    "air_disgenet_gbd_dict['Major Depressive Disorder'] = \"Major depressive disorder\"\n",
    "air_disgenet_gbd_dict['Malignant neoplasm of prostate'] = \"Prostate cancer\"\n",
    "air_disgenet_gbd_dict['Schizophrenia'] = \"Schizophrenia\"\n",
    "air_disgenet_gbd_dict['Endometrial Carcinoma'] = \"Uterine cancer\"\n",
    "air_disgenet_gbd_dict['Esophageal Neoplasms'] = 'Esophageal cancer'\n",
    "air_disgenet_gbd_dict['Mammary Neoplasms, Experimental'] = 'Breast cancer'\n",
    "air_disgenet_gbd_dict['Stomach Neoplasms'] = \"Stomach cancer\"\n",
    "air_disgenet_gbd_dict['Diabetes Mellitus, Non-Insulin-Dependent'] = \"Diabetes mellitus type 2\"\n",
    "air_disgenet_gbd_dict['Colorectal Neoplasms'] = 'Colon and rectum cancer'\n",
    "air_disgenet_gbd_dict['Asthma'] = 'Asthma'\n",
    "air_disgenet_gbd_dict['Liver carcinoma'] = 'Liver cancer'\n",
    "air_disgenet_gbd_dict['Rheumatoid Arthritis'] = \"Rheumatoid arthritis\"\n",
    "air_disgenet_gbd_dict['Breast Carcinoma'] = 'Breast cancer'\n",
    "air_disgenet_gbd_dict['Malignant neoplasm of breast'] = 'Breast cancer'\n",
    "air_disgenet_gbd_dict['Pancreatic Neoplasm'] = \"Pancreatic cancer\"\n",
    "air_disgenet_gbd_dict['Malignant tumor of colon'] = 'Colon and rectum cancer'\n",
    "air_disgenet_gbd_dict['Lung Neoplasms'] = \"Tracheal, bronchus, and lung cancer\"\n",
    "air_disgenet_gbd_dict['Brain Ischemia'] = 'Ischemic stroke'\n",
    "air_disgenet_gbd_dict['Carcinoma, Basal Cell'] = 'Non-melanoma skin cancer (basal-cell carcinoma)'\n",
    "air_disgenet_gbd_dict['Lymphoma, Non-Hodgkin'] = \"Non-Hodgkin lymphoma\"\n",
    "air_disgenet_gbd_dict['Diabetic Nephropathy'] = 'Diabetes and kidney diseases'\n",
    "air_disgenet_gbd_dict['Coronary heart disease'] = 'Ischemic heart disease'\n",
    "air_disgenet_gbd_dict['Kidney Failure, Chronic'] = 'Chronic kidney disease'\n",
    "air_disgenet_gbd_dict['Parkinson Disease'] = \"Parkinson's disease\"\n",
    "air_disgenet_gbd_dict['Autistic Disorder'] = \"Autism spectrum disorders\"\n",
    "air_disgenet_gbd_dict['Pancreatic carcinoma'] = \"Pancreatic cancer\"\n",
    "air_disgenet_gbd_dict['Breast adenocarcinoma'] = 'Breast cancer'\n",
    "air_disgenet_gbd_dict['Thyroid carcinoma'] = 'Thyroid cancer'\n",
    "air_disgenet_gbd_dict['Unipolar Depression'] = \"Major depressive disorder\"\n",
    "air_disgenet_gbd_dict['Polycystic Ovary Syndrome'] = \"Polycystic ovarian syndrome\"\n",
    "air_disgenet_gbd_dict['Squamous cell carcinoma of lung'] = \"Tracheal, bronchus, and lung cancer\"\n",
    "air_disgenet_gbd_dict['Glioblastoma Multiforme'] = \"Brain and central nervous system cancer\"\n",
    "air_disgenet_gbd_dict['Esophageal carcinoma'] = 'Esophageal cancer'\n",
    "air_disgenet_gbd_dict['Parkinsonian Disorders'] = \"Parkinson's disease\"\n",
    "air_disgenet_gbd_dict['Mental Depression'] = \"Major depressive disorder\"\n",
    "air_disgenet_gbd_dict['Coronary Artery Disease'] = 'Ischemic heart disease'\n",
    "air_disgenet_gbd_dict['Cleft Palate'] = \"Orofacial clefts\"\n",
    "air_disgenet_gbd_dict['Chronic Obstructive Airway Disease'] = 'Chronic obstructive pulmonary disease'\n",
    "air_disgenet_gbd_dict['Uterine Cervical Neoplasm'] = \"Uterine cancer\"\n",
    "air_disgenet_gbd_dict['Pulmonary Fibrosis'] = \"Interstitial lung disease and pulmonary sarcoidosis\"\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "a0b5527b-1b5b-40ae-8b8b-d9dbb1141cad",
   "metadata": {},
   "outputs": [],
   "source": [
    "country_selected_dis_mean_prox_dict = {}\n",
    "for country,disset in country_selected_dis_dict.items():\n",
    "    dis_dict = {}\n",
    "    for dis in disset:\n",
    "        z_list = [] \n",
    "        if len(country_selected_chem_id_dict[country])>0:\n",
    "            for chem in country_selected_chem_id_dict[country]:\n",
    "                try:\n",
    "                    z_list.append(real_significant_final_exp_proximity[chem,dis][1])\n",
    "                except:\n",
    "                    pass\n",
    "        dis_dict[dis]= np.round(np.mean(z_list),2)\n",
    "                    \n",
    "    country_selected_dis_mean_prox_dict[country] = dis_dict\n",
    "\n",
    "#Here, we have to clean the diseases that appears in each countries, considering only the ones that are reported in\n",
    "#GBD and using a little of manual curation for merging some terms (for example breast cancer and carcinoma of breast)\n",
    "#Now, we can create a country proximal diseases dictionary with the diseases taken from GBD\n",
    "\n",
    "\n",
    "country_selected_chem_id_gbd_dict = {}\n",
    "for country,dis_dict in country_selected_dis_mean_prox_dict.items():\n",
    "    if len(dis_dict)>0:\n",
    "        dis_z_list = []\n",
    "        for dis,z in dis_dict.items():\n",
    "            try:\n",
    "                dis_z_list.append([air_disgenet_gbd_dict[dis],z])\n",
    "            except:\n",
    "                pass\n",
    "        country_selected_chem_id_gbd_dict[country] = dis_z_list\n",
    "\n",
    "\n",
    "\n",
    "country_selected_chem_id_gbd_mean_dict = {}\n",
    "for country,dislist in country_selected_chem_id_gbd_dict.items():\n",
    "    dis_dict = {}\n",
    "    dis_list = []\n",
    "    for el in dislist:\n",
    "        dis_list.append(el[0])\n",
    "    dis_list_counter_dict = dict(Counter(dis_list).most_common())\n",
    "    for dis,count in dis_list_counter_dict.items():\n",
    "        z_list = []\n",
    "        if count>1:\n",
    "            for el in dislist:\n",
    "                if dis==el[0]:\n",
    "                    z_list.append(el[1])\n",
    "            dis_dict[dis]=np.round(np.mean(z_list),2)\n",
    "        else:\n",
    "            for el in dislist:\n",
    "                if dis==el[0]:\n",
    "                    dis_dict[el[0]] = el[1]\n",
    "    country_selected_chem_id_gbd_mean_dict[country]=dis_dict\n",
    "\n",
    "            "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "b6f8e10d-6a22-461e-9b9e-07084209c33c",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "cfe59d7d-7651-4ba5-b21f-765762ce19f8",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Let's import the distant diseases\n",
    "with open('output/exp_random/faraway_exp_dis_dict.pickle', 'rb') as handle:\n",
    "    faraway_exp_dis_dict = pk.load(handle)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "b72ad130-7634-4555-b8af-95b0c7c72f1b",
   "metadata": {},
   "outputs": [],
   "source": [
    "country_selected_faraway_dis_dict={}\n",
    "for country,chemset in country_selected_chem_id_dict.items():\n",
    "    if len(chemset)>0:\n",
    "        far_away_dis_set = []\n",
    "        for chem in chemset:\n",
    "            for chem_dis_pair in faraway_exp_dis_dict.keys():\n",
    "                if chem in chem_dis_pair[0]:\n",
    "                    far_away_dis_set.append(chem_dis_pair[1])\n",
    "        country_selected_faraway_dis_dict[country] = set(far_away_dis_set)\n",
    "\n",
    "faraway_air_disset = []\n",
    "for country,disset in country_selected_faraway_dis_dict.items():\n",
    "    for dis in disset:\n",
    "        if dis not in faraway_air_disset:\n",
    "            faraway_air_disset.append(dis)\n",
    "\n",
    "print(len(faraway_air_disset))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "48085409-da13-42dc-aa9a-2f2887466006",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Let's convert the name of the distant diseases\n",
    "\n",
    "long_faraway_air_disset_gbd_dict = {}\n",
    "for dis in faraway_air_disset:\n",
    "    dis_disgenet_higher = dis.higher()\n",
    "    for dis_gbd in disease_set:\n",
    "        dis_gbd_higher = dis_gbd.higher()\n",
    "        if len(get_overlap(dis_disgenet_higher,dis_gbd_higher))>0.7*len(dis_disgenet_higher):\n",
    "            long_faraway_air_disset_gbd_dict[dis] = dis_gbd\n",
    "\n",
    "print(len(long_faraway_air_disset_gbd_dict))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c902f433-1903-481a-acb7-41cae5baeded",
   "metadata": {},
   "outputs": [],
   "source": [
    "n_long_faraway_air_disset_gbd_dict = {}\n",
    "for dis1 in faraway_air_disset:\n",
    "    dis1_processed = preprocess_string(dis1)\n",
    "    # Initialize best match tracking\n",
    "    best_match = (None, 0)  # (Best match disease name, similarity score)\n",
    "    for dis2 in disease_set:\n",
    "        dis2_processed = preprocess_string(dis2)\n",
    "        similarity = get_similarity_score(dis1_processed, dis2_processed)\n",
    "        # Update if this is the best match so far\n",
    "        if similarity > best_match[1]:\n",
    "            best_match = (dis2, similarity)\n",
    "    \n",
    "    # You may want to set a threshold for \"good enough\" matching\n",
    "    if best_match[1] > 50:  # This threshold\n",
    "        n_long_faraway_air_disset_gbd_dict[dis1] = best_match[0]\n",
    "\n",
    "print(len(n_long_faraway_air_disset_gbd_dict))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "e5210df7-1a0b-443d-9d6a-b08240d4cfea",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Let's remove those diseases that had the wrong march\n",
    "\n",
    "\n",
    "keys_to_delete=['Weill-Marchesani-Like Syndrome','HYPOGONADOTROPIC HYPOGONADISM 11 WITH OR WITHOUT ANOSMIA','Night blindness, congenital stationary',\n",
    "             'Narcolepsy 1', 'DEAFNESS, X-LINKED 4 (disorder)','Juvenile nephropathic cystinosis (disorder)', 'Congenital Pain Insensitivity',\n",
    "                'Nephropathic cystinosis','Congenital long QT syndrome','Glomerulocystic Kidney Disease with Hyperuricemia and Isosthenuria',\n",
    "                'Keutel syndrome','FICOLIN 3 DEFICIENCY','Autosomal Dominant Nocturnal Frontal Lobe Epilepsy','Cavitary Optic Disc Anomalies',\n",
    "                'alpha, alpha-Trehalase deficiency','Nephritis, Tubulointerstitial','Cerebral ventriculomegaly','HYPOGONADOTROPIC HYPOGONADISM 4 WITH OR WITHOUT ANOSMIA',\n",
    "                'Goiter','Medullary Cystic Kidney Disease Type 2','Malabsorption Syndrome','CATARACT 39, MULTIPLE TYPES',\n",
    "                'Long Qt Syndrome 10','RETINITIS PIGMENTOSA 49','Olmsted syndrome','Cystinosis','HIRSCHSPRUNG DISEASE, SUSCEPTIBILITY TO, 4',\n",
    "                'MASP2 Deficiency','Cone monochromatism','Enterokinase Deficiency','Retinitis Pigmentosa 7','Narcolepsy','Jarcho-Levin syndrome',\n",
    "                'Waardenburg Syndrome, Type 4b','NIGHT BLINDNESS, CONGENITAL STATIONARY, TYPE 1E','Factor XIII, B Subunit, Deficiency Of',\n",
    "                'Renal hypouricemia','MYASTHENIC SYNDROME, CONGENITAL, 1A, Shigh-CHANNEL','BLEEDING DISORDER, PLATELET-TYPE, 20',\n",
    "                'SMALL PATELLA SYNDROME','Amish Infantile Epilepsy Syndrome','Scoliosis, unspecified','Focal glomerulosclerosis',\n",
    "                'Meningococcal Infections','Indifference to Pain, Congenital, Autosomal Recessive','Neuropathy, Distal Hereditary Motor, Type VIIA',\n",
    "                'Obstruction of pelviureteric junction','TELANGIECTASIA, HEREDITARY HEMORRHAGIC, TYPE 5','Marshall syndrome','STICKLER SYNDROME, TYPE II (disorder)',\n",
    "                'Histiocytosis','Hypopigmentation disorder','PAROXYSMAL EXTREME PAIN DISORDER','NEMALINE MYOPATHY 10','DEAFNESS, AUTOSOMAL RECESSIVE 42 (disorder)',\n",
    "                'Nivelon Nivelon Mabille syndrome','HERMANSKY-PUDLAK SYNDROME 4','Juvenile hemochromatosis','HEMOCHROMATOSIS, TYPE 2A','USHER SYNDROME, TYPE IJ',\n",
    "                'CATARACT 47','Usher Syndrome, Type IF','Hypouricemia, Renal, 2','NIGHT BLINDNESS, CONGENITAL STATIONARY, TYPE 1G','Primary Erythermalgia',\n",
    "                'Trimethylaminuria','Hypertrichosis','Congenital disorder of glycosylation, type 2C','Night Blindness, Congenital Stationary, Autosomal Dominant 3',\n",
    "                'Lysinuric Protein Intolerance','RETINITIS PIGMENTOSA, JUVENILE, LRAT-RELATED','Trichotillomania','Hydrocephalus','Usher Syndrome, Type I',\n",
    "                'STARGARDT DISEASE 1 (disorder)','Perisylvian syndrome','Liver Failure','CATARACT 4, MULTIPLE TYPES','DEAFNESS, AUTOSOMAL RECESSIVE 1A (disorder)',\n",
    "                'CILIARY DYSKINESIA, PRIMARY, 7 (disorder)','Vesico-Ureteral Reflux','VAN DER WOUDE SYNDROME 1','Hyperaldosteronism','Popliteal pterygium syndrome',\n",
    "                'Riboflavin Deficiency','Zunich neuroectodermal syndrome','Hypercalcemia, Infantile','CONE-ROD DYSTROPHY 3 (disorder)','RETINITIS PIGMENTOSA 19','CONGENITAL DISORDER OF GLYCOSYLATION, TYPE Ip',\n",
    "                'Juvenile macular degeneration',\"Stargardt's disease\",'RETINITIS PIGMENTOSA 48'\n",
    "                \n",
    "\n",
    "]\n",
    "\n",
    "n_long_faraway_air_disset_gbd_dict_cleaned = {}\n",
    "for disgenet_dis, gbd_dis in n_long_faraway_air_disset_gbd_dict.items():\n",
    "    if disgenet_dis not in keys_to_delete:\n",
    "        n_long_faraway_air_disset_gbd_dict_cleaned[disgenet_dis] = gbd_dis\n",
    "\n",
    "n_long_faraway_air_disset_gbd_dict_cleaned['CATARACT 39, MULTIPLE TYPES'] = 'Cataract'\n",
    "n_long_faraway_air_disset_gbd_dict_cleaned['NIGHT BLINDNESS, CONGENITAL STATIONARY, TYPE 1E'] = 'Blindness and vision loss'\n",
    "n_long_faraway_air_disset_gbd_dict_cleaned['CATARACT 4, MULTIPLE TYPES'] = 'Cataract'\n",
    "n_long_faraway_air_disset_gbd_dict_cleaned['CATARACT 47'] = 'Cataract'\n",
    "\n",
    "faraway_dis_list = list(n_long_faraway_air_disset_gbd_dict_cleaned.values())\n",
    "faraway_dis_list.extend(['Congenital musculoskeletal and limb anomalies','Congenital birth defects','Congenital heart anomalies',\n",
    "                        'Turner syndrome','Klinefelter syndrome'])\n",
    "\n",
    "print(len(faraway_dis_list))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "a666a7eb-10b6-4702-859d-d3fd985683c9",
   "metadata": {},
   "outputs": [],
   "source": [
    "faraway_manually_added_diseases = ['Congenital musculoskeletal and limb anomalies','Congenital birth defects','Congenital heart anomalies']\n",
    "\n",
    "country_faraway_dis_dict = {}\n",
    "for country,dis_dict in country_selected_chem_id_gbd_mean_dict.items():\n",
    "    chem_set = country_selected_chem_id_dict[country]\n",
    "    country_faraway_dis_list = []\n",
    "    country_specific_faraway_dis_dict = {}\n",
    "    \n",
    "    for dis in faraway_dis_list:\n",
    "        if dis not in dis_dict.keys():\n",
    "            country_faraway_dis_list.append(dis)\n",
    "    for dis in country_faraway_dis_list:\n",
    "        if dis not in faraway_manually_added_diseases:\n",
    "            disgenet_name_list = []\n",
    "            z_score_list = []\n",
    "            for disgenet_name,gbd_id in n_long_faraway_air_disset_gbd_dict_cleaned.items():\n",
    "                if dis == gbd_id:\n",
    "                    disgenet_name_list.append(disgenet_name)\n",
    "                    for chem in chem_set:\n",
    "                        try:\n",
    "                            z_score_list.append(exp_proximity_not_significant[chem,disgenet_name][1])\n",
    "                        except:\n",
    "                            pass\n",
    "            if np.mean(z_score_list)>1.5:\n",
    "                country_specific_faraway_dis_dict[dis]=np.mean(z_score_list)\n",
    "        elif dis=='Congenital heart anomalies':\n",
    "            disgenet_name_list = ['Congenital Heart Defects','Deafness, Congenital Heart Defects, and Posterior Embryotoxon',\n",
    "                                 'Congenital long QT syndrome']\n",
    "            z_score_list = []\n",
    "            for disgenet_name in disgenet_name_list:\n",
    "                for chem in chem_set:\n",
    "                    try:\n",
    "                        z_score_list.append(exp_proximity_not_significant[chem,disgenet_name][1])\n",
    "                    except:\n",
    "                        pass\n",
    "            if np.mean(z_score_list)>1.5:\n",
    "                country_specific_faraway_dis_dict[dis]=np.mean(z_score_list)\n",
    "        \n",
    "        elif dis=='Congenital musculoskeletal and limb anomalies':\n",
    "            disgenet_name_list = ['Congenital Fiber Type Disproportion','Myopathy, Actin, Congenital, with Excess of Thin Myofilaments',\n",
    "                                 'Myopathy, Actin, Congenital, With Cores','Muscular Dystrophy, Congenital, Megaconial Type',\n",
    "                                 'Myotonia Congenita','Myasthenic Syndromes, Congenital','Muscular Dystrophy, Congenital, Megaconial Type',\n",
    "                                 'Congenital Myasthenic Syndromes, Postsynaptic','Congenital Myasthenic Syndromes, Presynaptic',\n",
    "                                 'Muscular Dystrophy, Congenital, Due To Integrin Alpha-7 Deficiency','Muscular Dystrophy, Congenital, Type 1D',\n",
    "                                 'Congenital Fibrosis of the Extraocular Muscles']\n",
    "            z_score_list = []\n",
    "            for disgenet_name in disgenet_name_list:\n",
    "                for chem in chem_set:\n",
    "                    try:\n",
    "                        z_score_list.append(exp_proximity_not_significant[chem,disgenet_name][1])\n",
    "                    except:\n",
    "                        pass\n",
    "            if np.mean(z_score_list)>1.5:\n",
    "                country_specific_faraway_dis_dict[dis]=np.mean(z_score_list)\n",
    "            \n",
    "        elif dis=='Congenital birth defects':\n",
    "            disgenet_name_list = [dis for dis in list(diseases_genes_associated_ppi.keys()) if 'Congenital' in dis or 'congenital' in dis]\n",
    "            z_score_list = []\n",
    "            for disgenet_name in disgenet_name_list:\n",
    "                for chem in chem_set:\n",
    "                    try:\n",
    "                        z_score_list.append(exp_proximity_not_significant[chem,disgenet_name][1])\n",
    "                    except:\n",
    "                        pass\n",
    "            if np.mean(z_score_list)>1.5:\n",
    "                country_specific_faraway_dis_dict[dis]=np.mean(z_score_list)\n",
    "\n",
    "        elif dis=='Turner syndrome':\n",
    "            disgenet_name_list = [dis for dis in list(diseases_genes_associated_ppi.keys()) if 'Turner' in dis]\n",
    "            z_score_list = []\n",
    "            for disgenet_name in disgenet_name_list:\n",
    "                for chem in chem_set:\n",
    "                    try:\n",
    "                        z_score_list.append(exp_proximity_not_significant[chem,disgenet_name][1])\n",
    "                    except:\n",
    "                        pass\n",
    "            if np.mean(z_score_list)>1.5:\n",
    "                country_specific_faraway_dis_dict[dis]=np.mean(z_score_list)\n",
    "    country_faraway_dis_dict[country] = country_specific_faraway_dis_dict\n",
    "\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "91e0c906-4b6e-4ec3-95cc-19405fc8eeec",
   "metadata": {},
   "outputs": [],
   "source": [
    "no_prox_air_disset_gbd_dict = {}\n",
    "for dis1 in diseases_genes_associated_ppi.keys():\n",
    "    if 'syndrome' not in dis1:\n",
    "        if 'Syndrome' not in dis1 :\n",
    "            dis1_processed = preprocess_string(dis1)\n",
    "            # Initialize best match tracking\n",
    "            best_match = (None, 0)  # (Best match disease name, similarity score)\n",
    "            for dis2 in disease_set:\n",
    "                dis2_processed = preprocess_string(dis2)\n",
    "                similarity = get_similarity_score(dis1_processed, dis2_processed)\n",
    "                # Update if this is the best match so far\n",
    "                if similarity > best_match[1]:\n",
    "                    best_match = (dis2, similarity)\n",
    "            \n",
    "            # You may want to set a threshold for \"good enough\" matching\n",
    "            if best_match[1] > 70:  # This threshold\n",
    "                no_prox_air_disset_gbd_dict[dis1] = best_match[0]\n",
    "\n",
    "print(len(no_prox_air_disset_gbd_dict))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "5bfaed3d-51d8-4807-962a-f305ac99cb6f",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "#Let's clean the dictionary removing those terms that are wrong by the systematic analysis\n",
    "elements_to_delete = ['CHOPS SYNDROME','ADULT SYNDROME','Leber Congenital Amaurosis','COWDEN SYNDROME 6','Canavan Disease',\n",
    "                      'Cystic Kidney Diseases','Cystic Fibrosis','IRAK4 Deficiency','NOONAN SYNDROME 10','Glucagonoma',\n",
    "                      'MAST SYNDROME','Cholelithiasis','ovarian neoplasm','MYD88 Deficiency','Urocanase deficiency',\n",
    "                      'LONG QT SYNDROME 5','Malignant mesothelioma','Neurodevelopmental Disorders','Prion Diseases',\n",
    "                      'Mood Disorders','L-FERRITIN DEFICIENCY','Malignant neoplasm of liver','T-Cell Prolymphocytic Leukemia',\n",
    "                      'VERHEIJ SYNDROME','Tyrosine Kinase 2 Deficiency','Malignant neoplasm of testis','Steatohepatitis',\n",
    "                      'MEESTER-LOEYS SYNDROME','Heart Diseases','Cystinosis','Cardiomyopathy, Dilated, 1g','5-oxoprolinase deficiency',\n",
    "                      'Disorder of eye','Heart valve disease','HYDROCEPHALUS, CONGENITAL, 3, WITH BRAIN ANOMALIES','Transcobalamin II deficiency',\n",
    "                      'NOONAN SYNDROME 9','Hodgkin Disease','Malignant neoplasm of kidney','Acute Promyelocytic Leukemia','NGLY1 deficiency',\n",
    "                      'Movement Disorders','MASP2 Deficiency','Eye Diseases, Hereditary','FICOLIN 3 DEFICIENCY','MEND SYNDROME',\n",
    "                      'Scrapie','Endometrial Neoplasms','Diverticular Diseases','Manic Disorder','Malignant neoplasm of pancreas',\n",
    "                      'Enterokinase Deficiency','Malignant neoplasm of brain','COWDEN SYNDROME 7','Pneumonitis','Deficiency of galactokinase',\n",
    "                      'Arthritis, Psoriatic','Memory Disorders','Properdin deficiency disease','TORG-WINCHESTER SYNDROME','Thyroid Neoplasm',\n",
    "                      'NOONAN SYNDROME 7','COWCHOCK SYNDROME','Ataxia with vitamin E deficiency','INFLAMMATORY SKIN AND BOWEL DISEASE, NEONATAL, 1',\n",
    "                      'Skin Neoplasms','Glioma','Uterine Neoplasms','ZAP70 deficiency','AU-KLINE SYNDROME','Uremia','Kidney Calculi',\n",
    "                      'Lung Diseases, Interstitial','KLEEFSTRA SYNDROME 1','Protein C Deficiency','Gastritis, Atrophic','Peritoneal Fibrosis',\n",
    "                      'Stomach Neoplasms','Chronic Neutrophilic Leukemia','Nodal rhythm disorder','Malignant neoplasm of breast',\n",
    "                      'RAINE SYNDROME','Malignant neoplasm of ovary','Mitochondrial Diseases','COCOON SYNDROME','Thymus Neoplasms',\n",
    "                      'Malignant neoplasm of mouth','SWEENEY-COX SYNDROME','Celiac Disease','TYLOSIS WITH ESOPHAGEAL CANCER',\n",
    "                      'MUNGAN SYNDROME','Deficiency of prolidase','Galactosemias','Malignant neoplasm of prostate','prostatitis',\n",
    "                      'NOONAN SYNDROME 8','Dystonia','Glutamine deficiency, congenital','Liver neoplasms','C9 Deficiency','Retinal Diseases',\n",
    "                      'LONG QT SYNDROME 3','Mouth Neoplasms','Deficiency of fructokinase','Brain Neoplasms','BAKER-GORDON SYNDROME',\n",
    "                      'NOONAN SYNDROME 3','MYOPIA 6 (disorder)','Malignant neoplasm of lung','MACULAR DEGENERATION, EARLY-ONSET',\n",
    "                      'HIV Infections','Neonatal diabetes mellitus','Colitis','Dihydropyrimidinase deficiency','TENORIO SYNDROME',\n",
    "                      'Herpes encephalitis','Malignant melanoma of iris','GATA2 Deficiency','Intestinal Pseudo-Obstruction',\n",
    "                      \"Dent's disease\",'HEART DEFECTS, CONGENITAL, AND OTHER CONGENITAL ANOMALIES','Malignant neoplasm of stomach',\n",
    "                      'Osteitis Deformans','Fibrinogen Deficiency','Lung diseases','Metabolic Bone Disorder','Sialidase deficiency',\n",
    "                      'Lung Neoplasms','Iron Metabolism Disorders','Graves Disease','Malignant Neoplasms','Riboflavin Deficiency','Migraine Disorders',\n",
    "                      'Crohn Disease'\n",
    "]\n",
    "\n",
    "no_prox_air_disset_gbd_dict_cleaned = {}\n",
    "\n",
    "for disgenet_name,gbd_name in no_prox_air_disset_gbd_dict.items():\n",
    "    if disgenet_name not in elements_to_delete:\n",
    "        no_prox_air_disset_gbd_dict_cleaned[disgenet_name] = gbd_name\n",
    "\n",
    "noprox_gbd_list = set(list(no_prox_air_disset_gbd_dict_cleaned.values()))\n",
    "\n",
    "no_prox_air_disset_gbd_dict_cleaned_inv = {}\n",
    "for gbd_dis in noprox_gbd_list:\n",
    "    disgenet_name_list = [] \n",
    "    for disg_id,gbd_ in no_prox_air_disset_gbd_dict_cleaned.items():\n",
    "        if gbd_dis==gbd_:\n",
    "            disgenet_name_list.append(disg_id)\n",
    "    no_prox_air_disset_gbd_dict_cleaned_inv[gbd_dis] = disgenet_name_list\n",
    "\n",
    "print(len(no_prox_air_disset_gbd_dict_cleaned_inv))\n",
    "            \n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "b256dc5a-eff2-4611-83aa-73a143dc64b4",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "country_noprox_dis_dict = {}\n",
    "for country,dis_dict in country_selected_chem_id_gbd_mean_dict.items():\n",
    "    n_dis = len(dis_dict)\n",
    "    chem_set = country_selected_chem_id_dict[country]\n",
    "    country_noprox_dis_list = []\n",
    "    country_specific_noprox_dis_dict = {}\n",
    "    \n",
    "    for gbd_dis,dislist in no_prox_air_disset_gbd_dict_cleaned_inv.items():\n",
    "        if gbd_dis not in dis_dict.keys():\n",
    "            if gbd_dis not in list(country_faraway_dis_dict[country].keys()):\n",
    "                country_noprox_dis_list.append(gbd_dis)\n",
    "    for dis in country_noprox_dis_list:\n",
    "        disgenet_name_list = no_prox_air_disset_gbd_dict_cleaned_inv[dis]\n",
    "        z_score_list = []\n",
    "        for chem in chem_set:\n",
    "            for disgenet_name in disgenet_name_list:\n",
    "                try:\n",
    "                    z_score_list.append(exp_proximity_not_significant[chem,disgenet_name][1])\n",
    "                except:\n",
    "                    pass\n",
    "        if np.mean(z_score_list)>-1.5:\n",
    "            country_specific_noprox_dis_dict[dis]=np.mean(z_score_list)\n",
    "    country_specific_noprox_dis_dict_sort = {k: v for k, v in sorted(country_specific_noprox_dis_dict.items(), key=lambda item: item[1])}\n",
    "    \n",
    "    if len(country_specific_noprox_dis_dict_sort)>n_dis:\n",
    "        country_specific_noprox_dis_dict_sort_keys = list(country_specific_noprox_dis_dict_sort.keys())\n",
    "        n_country_specific_noprox_dis_dict_sort = {}\n",
    "        for k in list(country_specific_noprox_dis_dict_sort_keys)[:n_dis]:\n",
    "            n_country_specific_noprox_dis_dict_sort[k]=country_specific_noprox_dis_dict_sort[k]\n",
    "        \n",
    "        country_noprox_dis_dict[country] = n_country_specific_noprox_dis_dict_sort\n",
    "    else:\n",
    "        country_noprox_dis_dict[country] = country_specific_noprox_dis_dict_sort\n",
    " "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "1ce7b51e-0e30-466d-ae9e-e7826ef8e501",
   "metadata": {},
   "outputs": [],
   "source": [
    "for k,v in country_noprox_dis_dict.items():\n",
    "    print(k,len(v))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "b7e536f7-c232-45c7-abb9-77f267a9e3d9",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Let's import the incidence data\n",
    "with open('output/GBD/country_dis_incidence.pickle', 'rb') as handle:\n",
    "    country_dis_incidence = pk.load(handle)\n",
    "    \n",
    "with open('output/GBD/country_dis_prevalence.pickle', 'rb') as handle:\n",
    "    country_dis_prevalence = pk.load(handle)\n",
    "    \n",
    "with open('output/GBD/country_dis_DALYs.pickle', 'rb') as handle:\n",
    "    country_dis_DALYs = pk.load(handle)\n",
    "    \n",
    "with open('output/GBD/country_dis_mortality.pickle', 'rb') as handle:\n",
    "    country_dis_mortality = pk.load(handle)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "f6f2acea-9bc7-4105-a4a1-7a95bb98ab1d",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Let's just merge the no-proximal dictionary and the faraway dictionary\n",
    "new_country_distant_dis_dict = {}\n",
    "for country,dis_dict in country_selected_chem_id_gbd_mean_dict.items():\n",
    "    n_proxdis = len(dis_dict)\n",
    "    n_distantdis = len(country_faraway_dis_dict[country])\n",
    "    noprox_list = list(country_noprox_dis_dict[country].keys())\n",
    "    missing_dis = n_proxdis-n_distantdis\n",
    "    try:\n",
    "        dislist = noprox_list[:missing_dis]\n",
    "        distant_dict = country_faraway_dis_dict[country].copy()\n",
    "        for dis in dislist:\n",
    "            distant_dict[dis] = country_noprox_dis_dict[country][dis]\n",
    "        new_country_distant_dis_dict[country] = distant_dict\n",
    "    except:\n",
    "        distant_dict = {**country_faraway_dis_dict[country], **country_noprox_dis_dict[country]}\n",
    "        new_country_distant_dis_dict[country] = distant_dict"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "eb7f0579-8816-4cdf-b940-cfa4db3a4b4f",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "63bd0405-167d-4bd8-be6e-9761b844b0ca",
   "metadata": {},
   "outputs": [],
   "source": [
    "proximal_dis_list=[] \n",
    "for country, disdict in country_selected_chem_id_gbd_mean_dict.items():\n",
    "    proximal_dis_list.extend(list(disdict.keys()))\n",
    "    \n",
    "proximal_dis_incidence={}\n",
    "country_set=set(GBD_df['location'])\n",
    "for dis in set(proximal_dis_list):\n",
    "    incidence_list=[]\n",
    "    for country in country_set:\n",
    "        incidence_list.append(country_dis_incidence[country][dis])\n",
    "    proximal_dis_incidence[dis]=incidence_list\n",
    "\n",
    "\n",
    "proximal_dis_prevalence={}\n",
    "for dis in set(proximal_dis_list):\n",
    "    prevalence_list=[]\n",
    "    for country in country_set:\n",
    "        prevalence_list.append(country_dis_prevalence[country][dis])\n",
    "    proximal_dis_prevalence[dis]=prevalence_list\n",
    "\n",
    "proximal_dis_mortality={}\n",
    "for dis in set(proximal_dis_list):\n",
    "    mortality_list=[]\n",
    "    for country in country_set:\n",
    "        mortality_list.append(country_dis_mortality[country][dis])\n",
    "    proximal_dis_mortality[dis]=mortality_list\n",
    "\n",
    "proximal_dis_daly={}\n",
    "for dis in set(proximal_dis_list):\n",
    "    daly_list=[]\n",
    "    for country in country_set:\n",
    "        daly_list.append(country_dis_DALYs[country][dis])\n",
    "    proximal_dis_daly[dis]=daly_list"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "318af031-cd15-4ee1-8dfe-0dd0a96470b3",
   "metadata": {},
   "outputs": [],
   "source": [
    "import seaborn as sns\n",
    "country_proximal_zscore_dict={}\n",
    "for country,dis_dict in country_selected_chem_id_gbd_mean_dict.items():\n",
    "    country_proximal_zscore_list=[]\n",
    "    for dis in list(dis_dict.keys()):\n",
    "        mu = np.mean(proximal_dis_incidence[dis])\n",
    "        std = np.std(proximal_dis_incidence[dis])\n",
    "\n",
    "        z = (country_dis_incidence[country][dis]-mu)/std\n",
    "        country_proximal_zscore_list.append(z)\n",
    "\n",
    "        from scipy.stats import norm\n",
    "        ptemp = norm.sf(abs(z))\n",
    "    country_proximal_zscore_dict[country]=country_proximal_zscore_list"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "8c082fee-5e1b-44dc-bcd6-0613ef235924",
   "metadata": {},
   "outputs": [],
   "source": [
    "distant_dis_list=[] \n",
    "for country, disdict in new_country_distant_dis_dict.items():\n",
    "    distant_dis_list.extend(list(disdict.keys()))\n",
    "    \n",
    "distant_dis_incidence={}\n",
    "country_set=set(GBD_df['location'])\n",
    "for dis in set(distant_dis_list):\n",
    "    incidence_list=[]\n",
    "    for country in country_set:\n",
    "        incidence_list.append(country_dis_incidence[country][dis])\n",
    "    distant_dis_incidence[dis]=incidence_list\n",
    "\n",
    "distant_dis_prevalence={}\n",
    "for dis in set(distant_dis_list):\n",
    "    prevalence_list=[]\n",
    "    for country in country_set:\n",
    "        prevalence_list.append(country_dis_prevalence[country][dis])\n",
    "    distant_dis_prevalence[dis]=prevalence_list\n",
    "\n",
    "distant_dis_mortality={}\n",
    "for dis in set(distant_dis_list):\n",
    "    mortality_list=[]\n",
    "    for country in country_set:\n",
    "        mortality_list.append(country_dis_mortality[country][dis])\n",
    "    distant_dis_mortality[dis]=mortality_list\n",
    "\n",
    "distant_dis_daly={}\n",
    "for dis in set(distant_dis_list):\n",
    "    daly_list=[]\n",
    "    for country in country_set:\n",
    "        daly_list.append(country_dis_DALYs[country][dis])\n",
    "    distant_dis_daly[dis]=daly_list"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "1c33b4cd-5b8c-46a7-b021-238f9d1f8865",
   "metadata": {},
   "outputs": [],
   "source": [
    "air_country_distant_zscore_dict={}\n",
    "\n",
    "for country,dis_dict in new_country_distant_dis_dict.items():\n",
    "    country_distant_zscore_list=[]\n",
    "    for dis in list(dis_dict.keys()):\n",
    "        mu = np.mean(distant_dis_incidence[dis])\n",
    "        std = np.std(distant_dis_incidence[dis])\n",
    "\n",
    "        z = (country_dis_incidence[country][dis]-mu)/std\n",
    "        country_distant_zscore_list.append(z)\n",
    "\n",
    "        from scipy.stats import norm\n",
    "        ptemp = norm.sf(abs(z))\n",
    "    air_country_distant_zscore_dict[country]=country_distant_zscore_list"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "16cd867d-266c-4740-bd6e-a6eec46f1e72",
   "metadata": {},
   "outputs": [],
   "source": [
    "air_pollution_zscore_df = pd.DataFrame()\n",
    "\n",
    "selected_countries=set(air_country_distant_zscore_dict.keys())\n",
    "tot_country_list=[]\n",
    "tot_zscore_list=[]\n",
    "location_network_list=[]\n",
    "for country in selected_countries:\n",
    "    farawaylist_cleaned = [x for x in air_country_distant_zscore_dict[country] if str(x) != 'nan']\n",
    "    proximallist_cleaned = [x for x in country_proximal_zscore_dict[country] if str(x) != 'nan']\n",
    "    tot_zscorelist_partial=farawaylist_cleaned+proximallist_cleaned\n",
    "    for i in proximallist_cleaned:\n",
    "        location_network_list.append(\"proximal\")\n",
    "        tot_zscore_list.append(i)\n",
    "        tot_country_list.append(country)\n",
    "    for i in farawaylist_cleaned:\n",
    "        location_network_list.append(\"distant\")\n",
    "        tot_zscore_list.append(i)\n",
    "        tot_country_list.append(country)\n",
    "\n",
    "air_pollution_zscore_df['location']=tot_country_list\n",
    "air_pollution_zscore_df['z-score']=tot_zscore_list\n",
    "air_pollution_zscore_df['network_location']=location_network_list"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "070b0bcc-172b-4a64-af7d-1100f610b89c",
   "metadata": {},
   "outputs": [],
   "source": [
    "my_order = sorted(list(set(air_pollution_zscore_df['location'])))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "8a4576cc-12d2-438e-9f68-772ff7c0adb2",
   "metadata": {},
   "outputs": [],
   "source": [
    "import seaborn as sns\n",
    "fig, ax = plt.subplots(figsize=(15, 20))\n",
    "sns.set(style=\"white\")\n",
    "# Draw a nested boxplot to show bills by day and time\n",
    "g=sns.boxplot(x=\"z-score\", y=\"location\",orient=\"h\",\n",
    "            hue=\"network_location\", palette=[\"r\",\"b\"],order=my_order,\n",
    "            data=air_pollution_zscore_df)\n",
    "fontsize = 36\n",
    "plt.legend(loc=\"upper right\", frameon=True, fontsize=fontsize)\n",
    "plt.yticks(fontsize=36)\n",
    "#g.set_xscale(\"log\")\n",
    "for ax in plt.gcf().axes:\n",
    "    l = ax.get_xlabel()\n",
    "    ax.set_xlabel(l, fontsize=36)\n",
    "    ax.set_ylabel(l, fontsize=36)\n",
    "#plt.title('Incidence of network proximal and distant diseases to air polluted from 2000 to 2019', fontsize=25)\n",
    "plt.ylabel('', fontsize=25)\n",
    "plt.xlabel('', fontsize=25)\n",
    "ax.spines['top'].set_visible(False)\n",
    "ax.spines['right'].set_visible(False)\n",
    "ax.spines['bottom'].set_visible(True)\n",
    "ax.spines['left'].set_visible(True)\n",
    "#ax.set_yticklabels(ax.get_yticks(), size = 36)\n",
    "ax.set_xticklabels(ax.get_xticks(), size = 36)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "7e163b79-d41d-4a9f-8171-c78a4c26bad6",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Let's test the difference between proximal and no-proximal (because with distant is kind of obvious and it doesn't\n",
    "#make so much sense)\n",
    "import scipy\n",
    "for country in country_proximal_zscore_dict.keys():\n",
    "    try:\n",
    "        proximallist_cleaned = [x for x in country_proximal_zscore_dict[country] if str(x) != 'nan']\n",
    "        distantlist_cleaned = [x for x in air_country_distant_zscore_dict[country] if str(x) != 'nan']\n",
    "        prox_mean = np.mean(proximallist_cleaned)\n",
    "        distant_mean = np.mean(distantlist_cleaned)\n",
    "        pval_noprox = scipy.stats.mannwhitneyu(proximallist_cleaned, distantlist_cleaned, use_continuity=True, alternative='greater')[1]\n",
    "        print(country,prox_mean,distant_mean,round(pval_noprox,4),round(pval_distant,4))\n",
    "    except:\n",
    "        print(country,prox_mean,distant_mean,round(pval_noprox,4))\n",
    "        "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "2e111f55-391b-43b8-b655-a0ee3433b6dd",
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "\n",
    "fig, ax = plt.subplots(figsize=(15, 20))\n",
    "sns.set(style=\"white\")\n",
    "g = sns.boxplot(x=\"z-score\", y=\"location\", orient=\"h\",\n",
    "                hue=\"network_location\", palette=[\"r\", \"b\"], order=my_order,\n",
    "                data=air_pollution_zscore_df)\n",
    "fontsize = 36\n",
    "plt.legend(loc=\"upper right\", frameon=True, fontsize=fontsize)\n",
    "plt.yticks(fontsize=36)\n",
    "\n",
    "for ax in plt.gcf().axes:\n",
    "    l = ax.get_xlabel()\n",
    "    ax.set_xlabel(l, fontsize=36)\n",
    "    ax.set_ylabel(l, fontsize=36)\n",
    "\n",
    "plt.ylabel('', fontsize=25)\n",
    "plt.xlabel('', fontsize=25)\n",
    "ax.spines['top'].set_visible(False)\n",
    "ax.spines['right'].set_visible(False)\n",
    "ax.spines['bottom'].set_visible(True)\n",
    "ax.spines['left'].set_visible(True)\n",
    "ax.set_xticklabels(ax.get_xticks(), size=36)\n",
    "\n",
    "pv_list = []  \n",
    "locations = []  \n",
    "stars = []  \n",
    "\n",
    "for country in my_order:\n",
    "    proximallist_cleaned = [x for x in country_proximal_zscore_dict[country] if str(x) != 'nan']\n",
    "    distantlist_cleaned = [x for x in air_country_distant_zscore_dict[country] if str(x) != 'nan']\n",
    "    prox_mean = np.mean(proximallist_cleaned)\n",
    "    distant_mean = np.mean(distantlist_cleaned)\n",
    "    pval_noprox = scipy.stats.mannwhitneyu(proximallist_cleaned, distantlist_cleaned, use_continuity=True, alternative='greater')[1]\n",
    "    pv_list.append(pval_noprox)\n",
    "    locations.append(country)\n",
    "    if pval_noprox<0.001:\n",
    "        stars.append(\"***\")\n",
    "    elif pval_noprox>0.001 and pval_noprox<0.01:\n",
    "        stars.append(\"**\")\n",
    "    elif pval_noprox>0.01 and pval_noprox<0.05:\n",
    "        stars.append(\"*\")\n",
    "    else:\n",
    "        stars.append(\"n.s.\")\n",
    "\n",
    "for pv, loc, star in zip(pv_list, locations, stars):\n",
    "    yloc = my_order.index(loc) + 0.4 \n",
    "    xloc = max(air_pollution_zscore_df[air_pollution_zscore_df['location'] == loc]['z-score']) + 0.5  \n",
    "    ax.text(xloc, yloc, star, fontsize=fontsize, color=\"black\", ha='center')\n",
    "\n",
    "\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "3e18b2f4-bc71-4262-862d-f7d2790aed41",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "10492df9-fef2-4565-b2aa-6525be23e8b9",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Prevalence\n",
    "\n",
    "country_proximal_zscore_dict={}\n",
    "for country,dis_dict in country_selected_chem_id_gbd_mean_dict.items():\n",
    "    country_proximal_zscore_list=[]\n",
    "    for dis in list(dis_dict.keys()):\n",
    "        mu = np.mean(proximal_dis_prevalence[dis])\n",
    "        std = np.std(proximal_dis_prevalence[dis])\n",
    "        z = (country_dis_prevalence[country][dis]-mu)/std\n",
    "        country_proximal_zscore_list.append(z)\n",
    "        from scipy.stats import norm\n",
    "        ptemp = norm.sf(abs(z))\n",
    "    country_proximal_zscore_dict[country]=country_proximal_zscore_list\n",
    "\n",
    "air_country_distant_zscore_dict={}\n",
    "for country,dis_dict in new_country_distant_dis_dict.items():\n",
    "    country_distant_zscore_list=[]\n",
    "    for dis in list(dis_dict.keys()):\n",
    "        mu = np.mean(distant_dis_prevalence[dis])\n",
    "        std = np.std(distant_dis_prevalence[dis])\n",
    "        z = (country_dis_prevalence[country][dis]-mu)/std\n",
    "        country_distant_zscore_list.append(z)\n",
    "        from scipy.stats import norm\n",
    "        ptemp = norm.sf(abs(z))\n",
    "    air_country_distant_zscore_dict[country]=country_distant_zscore_list\n",
    "\n",
    "air_pollution_zscore_df = pd.DataFrame()\n",
    "selected_countries=set(air_country_distant_zscore_dict.keys())\n",
    "tot_country_list=[]\n",
    "tot_zscore_list=[]\n",
    "location_network_list=[]\n",
    "for country in selected_countries:\n",
    "    farawaylist_cleaned = [x for x in air_country_distant_zscore_dict[country] if str(x) != 'nan']\n",
    "    proximallist_cleaned = [x for x in country_proximal_zscore_dict[country] if str(x) != 'nan']\n",
    "    tot_zscorelist_partial=farawaylist_cleaned+proximallist_cleaned\n",
    "    for i in proximallist_cleaned:\n",
    "        location_network_list.append(\"proximal\")\n",
    "        tot_zscore_list.append(i)\n",
    "        tot_country_list.append(country)\n",
    "    for i in farawaylist_cleaned:\n",
    "        location_network_list.append(\"distant\")\n",
    "        tot_zscore_list.append(i)\n",
    "        tot_country_list.append(country)\n",
    "\n",
    "air_pollution_zscore_df['location']=tot_country_list\n",
    "air_pollution_zscore_df['z-score']=tot_zscore_list\n",
    "air_pollution_zscore_df['network_location']=location_network_list\n",
    "\n",
    "my_order = sorted(list(set(air_pollution_zscore_df['location'])))\n",
    "\n",
    "for country in country_proximal_zscore_dict.keys():\n",
    "    try:\n",
    "        proximallist_cleaned = [x for x in country_proximal_zscore_dict[country] if str(x) != 'nan']\n",
    "        distantlist_cleaned = [x for x in air_country_distant_zscore_dict[country] if str(x) != 'nan']\n",
    "        prox_mean = np.mean(proximallist_cleaned)\n",
    "        distant_mean = np.mean(distantlist_cleaned)\n",
    "        pval_noprox = scipy.stats.mannwhitneyu(proximallist_cleaned, distantlist_cleaned, use_continuity=True, alternative='greater')[1]\n",
    "        print(country,prox_mean,distant_mean,round(pval_noprox,4),round(pval_distant,4))\n",
    "    except:\n",
    "        print(country,prox_mean,distant_mean,round(pval_noprox,4))\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "281e8384-d53c-49c0-8c92-ed0d411f750d",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "fig, ax = plt.subplots(figsize=(15, 20))\n",
    "sns.set(style=\"white\")\n",
    "g = sns.boxplot(x=\"z-score\", y=\"location\", orient=\"h\",\n",
    "                hue=\"network_location\", palette=[\"#2B7A98\", \"#54CAF2\"], order=my_order,\n",
    "                data=air_pollution_zscore_df)\n",
    "fontsize = 36\n",
    "plt.legend(loc=\"upper right\", frameon=True, fontsize=fontsize)\n",
    "plt.yticks(fontsize=36)\n",
    "\n",
    "for ax in plt.gcf().axes:\n",
    "    l = ax.get_xlabel()\n",
    "    ax.set_xlabel(l, fontsize=36)\n",
    "    ax.set_ylabel(l, fontsize=36)\n",
    "\n",
    "plt.ylabel('', fontsize=25)\n",
    "plt.xlabel('', fontsize=25)\n",
    "ax.spines['top'].set_visible(False)\n",
    "ax.spines['right'].set_visible(False)\n",
    "ax.spines['bottom'].set_visible(True)\n",
    "ax.spines['left'].set_visible(True)\n",
    "ax.set_xticklabels(ax.get_xticks(), size=36)\n",
    "\n",
    "pv_list = []  \n",
    "locations = []  \n",
    "stars = []  \n",
    "\n",
    "for country in my_order:\n",
    "    proximallist_cleaned = [x for x in country_proximal_zscore_dict[country] if str(x) != 'nan']\n",
    "    distantlist_cleaned = [x for x in air_country_distant_zscore_dict[country] if str(x) != 'nan']\n",
    "    prox_mean = np.mean(proximallist_cleaned)\n",
    "    distant_mean = np.mean(distantlist_cleaned)\n",
    "    pval_noprox = scipy.stats.mannwhitneyu(proximallist_cleaned, distantlist_cleaned, use_continuity=True, alternative='greater')[1]\n",
    "    pv_list.append(pval_noprox)\n",
    "    locations.append(country)\n",
    "    if pval_noprox<0.001:\n",
    "        stars.append(\"***\")\n",
    "    elif pval_noprox>0.001 and pval_noprox<0.01:\n",
    "        stars.append(\"**\")\n",
    "    elif pval_noprox>0.01 and pval_noprox<0.05:\n",
    "        stars.append(\"*\")\n",
    "    else:\n",
    "        stars.append(\"n.s.\")\n",
    "\n",
    "for pv, loc, star in zip(pv_list, locations, stars):\n",
    "    yloc = my_order.index(loc) + 0.4 \n",
    "    xloc = max(air_pollution_zscore_df[air_pollution_zscore_df['location'] == loc]['z-score']) + 0.5  \n",
    "    ax.text(xloc, yloc, star, fontsize=fontsize, color=\"black\", ha='center')\n",
    "\n",
    "plt.savefig('output/Figures/SF12A_v01_post_rev.pdf',format='pdf',dpi=300, bbox_inches = \"tight\")\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "a0d5b572-5d09-4d8a-ac7c-9a5b53ec7a2d",
   "metadata": {},
   "outputs": [],
   "source": [
    "#mortality\n",
    "\n",
    "country_proximal_zscore_dict={}\n",
    "for country,dis_dict in country_selected_chem_id_gbd_mean_dict.items():\n",
    "    country_proximal_zscore_list=[]\n",
    "    for dis in list(dis_dict.keys()):\n",
    "        mu = np.mean(proximal_dis_mortality[dis])\n",
    "        std = np.std(proximal_dis_mortality[dis])\n",
    "        z = (country_dis_mortality[country][dis]-mu)/std\n",
    "        country_proximal_zscore_list.append(z)\n",
    "        from scipy.stats import norm\n",
    "        ptemp = norm.sf(abs(z))\n",
    "    country_proximal_zscore_dict[country]=country_proximal_zscore_list\n",
    "\n",
    "air_country_distant_zscore_dict={}\n",
    "for country,dis_dict in new_country_distant_dis_dict.items():\n",
    "    country_distant_zscore_list=[]\n",
    "    for dis in list(dis_dict.keys()):\n",
    "        mu = np.mean(distant_dis_mortality[dis])\n",
    "        std = np.std(distant_dis_mortality[dis])\n",
    "        z = (country_dis_mortality[country][dis]-mu)/std\n",
    "        country_distant_zscore_list.append(z)\n",
    "        from scipy.stats import norm\n",
    "        ptemp = norm.sf(abs(z))\n",
    "    air_country_distant_zscore_dict[country]=country_distant_zscore_list\n",
    "\n",
    "air_pollution_zscore_df = pd.DataFrame()\n",
    "selected_countries=set(air_country_distant_zscore_dict.keys())\n",
    "tot_country_list=[]\n",
    "tot_zscore_list=[]\n",
    "location_network_list=[]\n",
    "for country in selected_countries:\n",
    "    farawaylist_cleaned = [x for x in air_country_distant_zscore_dict[country] if str(x) != 'nan']\n",
    "    proximallist_cleaned = [x for x in country_proximal_zscore_dict[country] if str(x) != 'nan']\n",
    "    tot_zscorelist_partial=farawaylist_cleaned+proximallist_cleaned\n",
    "    for i in proximallist_cleaned:\n",
    "        location_network_list.append(\"proximal\")\n",
    "        tot_zscore_list.append(i)\n",
    "        tot_country_list.append(country)\n",
    "    for i in farawaylist_cleaned:\n",
    "        location_network_list.append(\"distant\")\n",
    "        tot_zscore_list.append(i)\n",
    "        tot_country_list.append(country)\n",
    "\n",
    "air_pollution_zscore_df['location']=tot_country_list\n",
    "air_pollution_zscore_df['z-score']=tot_zscore_list\n",
    "air_pollution_zscore_df['network_location']=location_network_list\n",
    "\n",
    "my_order = sorted(list(set(air_pollution_zscore_df['location'])))\n",
    "\n",
    "for country in country_proximal_zscore_dict.keys():\n",
    "    try:\n",
    "        proximallist_cleaned = [x for x in country_proximal_zscore_dict[country] if str(x) != 'nan']\n",
    "        distantlist_cleaned = [x for x in air_country_distant_zscore_dict[country] if str(x) != 'nan']\n",
    "        prox_mean = np.mean(proximallist_cleaned)\n",
    "        distant_mean = np.mean(distantlist_cleaned)\n",
    "        pval_noprox = scipy.stats.mannwhitneyu(proximallist_cleaned, distantlist_cleaned, use_continuity=True, alternative='greater')[1]\n",
    "        print(country,prox_mean,distant_mean,round(pval_noprox,4),round(pval_distant,4))\n",
    "    except:\n",
    "        print(country,prox_mean,distant_mean,round(pval_noprox,4))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "645cb93c-e73e-4ec1-a87a-76cabe6c3bb7",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "fig, ax = plt.subplots(figsize=(15, 20))\n",
    "sns.set(style=\"white\")\n",
    "g = sns.boxplot(x=\"z-score\", y=\"location\", orient=\"h\",\n",
    "                hue=\"network_location\", palette=[\"#2B7A98\", \"#54CAF2\"], order=my_order,\n",
    "                data=air_pollution_zscore_df)\n",
    "fontsize = 36\n",
    "plt.legend(loc=\"upper right\", frameon=True, fontsize=fontsize)\n",
    "plt.yticks(fontsize=36)\n",
    "\n",
    "for ax in plt.gcf().axes:\n",
    "    l = ax.get_xlabel()\n",
    "    ax.set_xlabel(l, fontsize=36)\n",
    "    ax.set_ylabel(l, fontsize=36)\n",
    "\n",
    "plt.ylabel('', fontsize=25)\n",
    "plt.xlabel('', fontsize=25)\n",
    "ax.spines['top'].set_visible(False)\n",
    "ax.spines['right'].set_visible(False)\n",
    "ax.spines['bottom'].set_visible(True)\n",
    "ax.spines['left'].set_visible(True)\n",
    "ax.set_xticklabels(ax.get_xticks(), size=36)\n",
    "\n",
    "pv_list = []  \n",
    "locations = []  \n",
    "stars = []  \n",
    "\n",
    "for country in my_order:\n",
    "    proximallist_cleaned = [x for x in country_proximal_zscore_dict[country] if str(x) != 'nan']\n",
    "    distantlist_cleaned = [x for x in air_country_distant_zscore_dict[country] if str(x) != 'nan']\n",
    "    prox_mean = np.mean(proximallist_cleaned)\n",
    "    distant_mean = np.mean(distantlist_cleaned)\n",
    "    pval_noprox = scipy.stats.mannwhitneyu(proximallist_cleaned, distantlist_cleaned, use_continuity=True, alternative='greater')[1]\n",
    "    pv_list.append(pval_noprox)\n",
    "    locations.append(country)\n",
    "    if pval_noprox<0.001:\n",
    "        stars.append(\"***\")\n",
    "    elif pval_noprox>0.001 and pval_noprox<0.01:\n",
    "        stars.append(\"**\")\n",
    "    elif pval_noprox>0.01 and pval_noprox<0.05:\n",
    "        stars.append(\"*\")\n",
    "    else:\n",
    "        stars.append(\"n.s.\")\n",
    "\n",
    "for pv, loc, star in zip(pv_list, locations, stars):\n",
    "    yloc = my_order.index(loc) + 0.4 \n",
    "    xloc = max(air_pollution_zscore_df[air_pollution_zscore_df['location'] == loc]['z-score']) + 0.5  \n",
    "    ax.text(xloc, yloc, star, fontsize=fontsize, color=\"black\", ha='center')\n",
    "\n",
    "plt.savefig('output/Figures/SF13A_v01_post_rev.pdf',format='pdf',dpi=300, bbox_inches = \"tight\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "046c7876-b4e8-4e61-b702-ce3a4dee0add",
   "metadata": {},
   "outputs": [],
   "source": [
    "#daly\n",
    "\n",
    "country_proximal_zscore_dict={}\n",
    "for country,dis_dict in country_selected_chem_id_gbd_mean_dict.items():\n",
    "    country_proximal_zscore_list=[]\n",
    "    for dis in list(dis_dict.keys()):\n",
    "        mu = np.mean(proximal_dis_daly[dis])\n",
    "        std = np.std(proximal_dis_daly[dis])\n",
    "        z = (country_dis_DALYs[country][dis]-mu)/std\n",
    "        country_proximal_zscore_list.append(z)\n",
    "        from scipy.stats import norm\n",
    "        ptemp = norm.sf(abs(z))\n",
    "    country_proximal_zscore_dict[country]=country_proximal_zscore_list\n",
    "\n",
    "air_country_distant_zscore_dict={}\n",
    "for country,dis_dict in new_country_distant_dis_dict.items():\n",
    "    country_distant_zscore_list=[]\n",
    "    for dis in list(dis_dict.keys()):\n",
    "        mu = np.mean(distant_dis_daly[dis])\n",
    "        std = np.std(distant_dis_daly[dis])\n",
    "        z = (country_dis_DALYs[country][dis]-mu)/std\n",
    "        country_distant_zscore_list.append(z)\n",
    "        from scipy.stats import norm\n",
    "        ptemp = norm.sf(abs(z))\n",
    "    air_country_distant_zscore_dict[country]=country_distant_zscore_list\n",
    "\n",
    "air_pollution_zscore_df = pd.DataFrame()\n",
    "selected_countries=set(air_country_distant_zscore_dict.keys())\n",
    "tot_country_list=[]\n",
    "tot_zscore_list=[]\n",
    "location_network_list=[]\n",
    "for country in selected_countries:\n",
    "    farawaylist_cleaned = [x for x in air_country_distant_zscore_dict[country] if str(x) != 'nan']\n",
    "    proximallist_cleaned = [x for x in country_proximal_zscore_dict[country] if str(x) != 'nan']\n",
    "    tot_zscorelist_partial=farawaylist_cleaned+proximallist_cleaned\n",
    "    for i in proximallist_cleaned:\n",
    "        location_network_list.append(\"proximal\")\n",
    "        tot_zscore_list.append(i)\n",
    "        tot_country_list.append(country)\n",
    "    for i in farawaylist_cleaned:\n",
    "        location_network_list.append(\"distant\")\n",
    "        tot_zscore_list.append(i)\n",
    "        tot_country_list.append(country)\n",
    "\n",
    "air_pollution_zscore_df['location']=tot_country_list\n",
    "air_pollution_zscore_df['z-score']=tot_zscore_list\n",
    "air_pollution_zscore_df['network_location']=location_network_list\n",
    "\n",
    "my_order = sorted(list(set(air_pollution_zscore_df['location'])))\n",
    "\n",
    "for country in country_proximal_zscore_dict.keys():\n",
    "    try:\n",
    "        proximallist_cleaned = [x for x in country_proximal_zscore_dict[country] if str(x) != 'nan']\n",
    "        distantlist_cleaned = [x for x in air_country_distant_zscore_dict[country] if str(x) != 'nan']\n",
    "        prox_mean = np.mean(proximallist_cleaned)\n",
    "        distant_mean = np.mean(distantlist_cleaned)\n",
    "        pval_noprox = scipy.stats.mannwhitneyu(proximallist_cleaned, distantlist_cleaned, use_continuity=True, alternative='greater')[1]\n",
    "        print(country,prox_mean,distant_mean,round(pval_noprox,4),round(pval_distant,4))\n",
    "    except:\n",
    "        print(country,prox_mean,distant_mean,round(pval_noprox,4))\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "f041f6a5-3217-49a4-9a3c-dd120ef362ab",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "fig, ax = plt.subplots(figsize=(15, 20))\n",
    "sns.set(style=\"white\")\n",
    "g = sns.boxplot(x=\"z-score\", y=\"location\", orient=\"h\",\n",
    "                hue=\"network_location\", palette=[\"#2B7A98\", \"#54CAF2\"], order=my_order,\n",
    "                data=air_pollution_zscore_df)\n",
    "fontsize = 36\n",
    "plt.legend(loc=\"upper right\", frameon=True, fontsize=fontsize)\n",
    "plt.yticks(fontsize=36)\n",
    "\n",
    "for ax in plt.gcf().axes:\n",
    "    l = ax.get_xlabel()\n",
    "    ax.set_xlabel(l, fontsize=36)\n",
    "    ax.set_ylabel(l, fontsize=36)\n",
    "\n",
    "plt.ylabel('', fontsize=25)\n",
    "plt.xlabel('', fontsize=25)\n",
    "ax.spines['top'].set_visible(False)\n",
    "ax.spines['right'].set_visible(False)\n",
    "ax.spines['bottom'].set_visible(True)\n",
    "ax.spines['left'].set_visible(True)\n",
    "ax.set_xticklabels(ax.get_xticks(), size=36)\n",
    "\n",
    "pv_list = []  \n",
    "locations = []  \n",
    "stars = []  \n",
    "\n",
    "for country in my_order:\n",
    "    proximallist_cleaned = [x for x in country_proximal_zscore_dict[country] if str(x) != 'nan']\n",
    "    distantlist_cleaned = [x for x in air_country_distant_zscore_dict[country] if str(x) != 'nan']\n",
    "    prox_mean = np.mean(proximallist_cleaned)\n",
    "    distant_mean = np.mean(distantlist_cleaned)\n",
    "    pval_noprox = scipy.stats.mannwhitneyu(proximallist_cleaned, distantlist_cleaned, use_continuity=True, alternative='greater')[1]\n",
    "    pv_list.append(pval_noprox)\n",
    "    locations.append(country)\n",
    "    if pval_noprox<0.001:\n",
    "        stars.append(\"***\")\n",
    "    elif pval_noprox>0.001 and pval_noprox<0.01:\n",
    "        stars.append(\"**\")\n",
    "    elif pval_noprox>0.01 and pval_noprox<0.05:\n",
    "        stars.append(\"*\")\n",
    "    else:\n",
    "        stars.append(\"n.s.\")\n",
    "\n",
    "for pv, loc, star in zip(pv_list, locations, stars):\n",
    "    yloc = my_order.index(loc) + 0.4 \n",
    "    xloc = max(air_pollution_zscore_df[air_pollution_zscore_df['location'] == loc]['z-score']) + 0.5  \n",
    "    ax.text(xloc, yloc, star, fontsize=fontsize, color=\"black\", ha='center')\n",
    "\n",
    "plt.savefig('output/Figures/SF14A_v01_post_rev.pdf',format='pdf',dpi=300, bbox_inches = \"tight\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d13a02db-e329-4949-bdcf-dcc0c1242f49",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "b1e82be1-fe1b-41dd-b10c-395758d4ba54",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "4fb21cf4-49aa-421a-83e9-88907c226ae3",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "a342148d-80bc-4374-a76c-90f59d0933fd",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Low\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "0c889a85-839c-49e7-a933-127fcd774e04",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Proximal\n",
    "low_country_selected_dis_mean_prox_dict = {}\n",
    "for country,disset in low_country_selected_dis_dict.items():\n",
    "    dis_dict = {}\n",
    "    for dis in disset:\n",
    "        z_list = [] \n",
    "        if len(low_country_selected_chem_id_dict[country])>0:\n",
    "            for chem in low_country_selected_chem_id_dict[country]:\n",
    "                try:\n",
    "                    z_list.append(real_significant_final_exp_proximity[chem,dis][1])\n",
    "                except:\n",
    "                    pass\n",
    "        dis_dict[dis]= np.round(np.mean(z_list),2)\n",
    "                    \n",
    "    low_country_selected_dis_mean_prox_dict[country] = dis_dict\n",
    "\n",
    "low_country_selected_chem_id_gbd_dict = {}\n",
    "for country,dis_dict in low_country_selected_dis_mean_prox_dict.items():\n",
    "    if len(dis_dict)>0:\n",
    "        dis_z_list = []\n",
    "        for dis,z in dis_dict.items():\n",
    "            try:\n",
    "                dis_z_list.append([air_disgenet_gbd_dict[dis],z])\n",
    "            except:\n",
    "                pass\n",
    "        low_country_selected_chem_id_gbd_dict[country] = dis_z_list\n",
    "\n",
    "low_country_selected_chem_id_gbd_mean_dict = {}\n",
    "for country,dislist in low_country_selected_chem_id_gbd_dict.items():\n",
    "    dis_dict = {}\n",
    "    dis_list = []\n",
    "    for el in dislist:\n",
    "        dis_list.append(el[0])\n",
    "    dis_list_counter_dict = dict(Counter(dis_list).most_common())\n",
    "    for dis,count in dis_list_counter_dict.items():\n",
    "        z_list = []\n",
    "        if count>1:\n",
    "            for el in dislist:\n",
    "                if dis==el[0]:\n",
    "                    z_list.append(el[1])\n",
    "            dis_dict[dis]=np.round(np.mean(z_list),2)\n",
    "        else:\n",
    "            for el in dislist:\n",
    "                if dis==el[0]:\n",
    "                    dis_dict[el[0]] = el[1]\n",
    "    low_country_selected_chem_id_gbd_mean_dict[country]=dis_dict\n",
    "\n",
    "proximal_dis_list=[] \n",
    "for country, disdict in low_country_selected_chem_id_gbd_mean_dict.items():\n",
    "    proximal_dis_list.extend(list(disdict.keys()))\n",
    "    \n",
    "proximal_dis_incidence={}\n",
    "country_set=set(GBD_df['location'])\n",
    "for dis in set(proximal_dis_list):\n",
    "    incidence_list=[]\n",
    "    for country in country_set:\n",
    "        incidence_list.append(country_dis_incidence[country][dis])\n",
    "    proximal_dis_incidence[dis]=incidence_list\n",
    "\n",
    "proximal_dis_prevalence={}\n",
    "for dis in set(proximal_dis_list):\n",
    "    prevalence_list=[]\n",
    "    for country in country_set:\n",
    "        prevalence_list.append(country_dis_prevalence[country][dis])\n",
    "    proximal_dis_prevalence[dis]=prevalence_list\n",
    "\n",
    "proximal_dis_mortality={}\n",
    "for dis in set(proximal_dis_list):\n",
    "    mortality_list=[]\n",
    "    for country in country_set:\n",
    "        mortality_list.append(country_dis_mortality[country][dis])\n",
    "    proximal_dis_mortality[dis]=mortality_list\n",
    "\n",
    "proximal_dis_daly={}\n",
    "for dis in set(proximal_dis_list):\n",
    "    daly_list=[]\n",
    "    for country in country_set:\n",
    "        daly_list.append(country_dis_DALYs[country][dis])\n",
    "    proximal_dis_daly[dis]=daly_list\n",
    "\n",
    "low_country_proximal_zscore_dict={}\n",
    "for country,dis_dict in low_country_selected_chem_id_gbd_mean_dict.items():\n",
    "    country_proximal_zscore_list=[]\n",
    "    for dis in list(dis_dict.keys()):\n",
    "        mu = np.mean(proximal_dis_incidence[dis])\n",
    "        std = np.std(proximal_dis_incidence[dis])\n",
    "\n",
    "        z = (country_dis_incidence[country][dis]-mu)/std\n",
    "        country_proximal_zscore_list.append(z)\n",
    "\n",
    "        from scipy.stats import norm\n",
    "        ptemp = norm.sf(abs(z))\n",
    "    if len(country_proximal_zscore_list)>0:\n",
    "        low_country_proximal_zscore_dict[country]=country_proximal_zscore_list\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "2aa9c893-64b5-43cc-be69-d0137ae8b952",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Distant\n",
    "low_country_selected_faraway_dis_dict={}\n",
    "for country,chemset in low_country_selected_chem_id_dict.items():\n",
    "    if len(chemset)>0:\n",
    "        far_away_dis_set = []\n",
    "        for chem in chemset:\n",
    "            for chem_dis_pair in faraway_exp_dis_dict.keys():\n",
    "                if chem in chem_dis_pair[0]:\n",
    "                    far_away_dis_set.append(chem_dis_pair[1])\n",
    "        low_country_selected_faraway_dis_dict[country] = set(far_away_dis_set)\n",
    "\n",
    "\n",
    "low_country_faraway_dis_dict = {}\n",
    "for country,dis_dict in low_country_selected_chem_id_gbd_mean_dict.items():\n",
    "    chem_set = low_country_selected_chem_id_dict[country]\n",
    "    country_faraway_dis_list = []\n",
    "    country_specific_faraway_dis_dict = {}\n",
    "    \n",
    "    for dis in faraway_dis_list:\n",
    "        if dis not in dis_dict.keys():\n",
    "            country_faraway_dis_list.append(dis)\n",
    "    for dis in country_faraway_dis_list:\n",
    "        if dis not in faraway_manually_added_diseases:\n",
    "            disgenet_name_list = []\n",
    "            z_score_list = []\n",
    "            for disgenet_name,gbd_id in n_long_faraway_air_disset_gbd_dict_cleaned.items():\n",
    "                if dis == gbd_id:\n",
    "                    disgenet_name_list.append(disgenet_name)\n",
    "                    for chem in chem_set:\n",
    "                        try:\n",
    "                            z_score_list.append(exp_proximity_not_significant[chem,disgenet_name][1])\n",
    "                        except:\n",
    "                            pass\n",
    "            if np.mean(z_score_list)>1.5:\n",
    "                country_specific_faraway_dis_dict[dis]=np.mean(z_score_list)\n",
    "        elif dis=='Congenital heart anomalies':\n",
    "            disgenet_name_list = ['Congenital Heart Defects','Deafness, Congenital Heart Defects, and Posterior Embryotoxon',\n",
    "                                 'Congenital long QT syndrome']\n",
    "            z_score_list = []\n",
    "            for disgenet_name in disgenet_name_list:\n",
    "                for chem in chem_set:\n",
    "                    try:\n",
    "                        z_score_list.append(exp_proximity_not_significant[chem,disgenet_name][1])\n",
    "                    except:\n",
    "                        pass\n",
    "            if np.mean(z_score_list)>1.5:\n",
    "                country_specific_faraway_dis_dict[dis]=np.mean(z_score_list)\n",
    "        \n",
    "        elif dis=='Congenital musculoskeletal and limb anomalies':\n",
    "            disgenet_name_list = ['Congenital Fiber Type Disproportion','Myopathy, Actin, Congenital, with Excess of Thin Myofilaments',\n",
    "                                 'Myopathy, Actin, Congenital, With Cores','Muscular Dystrophy, Congenital, Megaconial Type',\n",
    "                                 'Myotonia Congenita','Myasthenic Syndromes, Congenital','Muscular Dystrophy, Congenital, Megaconial Type',\n",
    "                                 'Congenital Myasthenic Syndromes, Postsynaptic','Congenital Myasthenic Syndromes, Presynaptic',\n",
    "                                 'Muscular Dystrophy, Congenital, Due To Integrin Alpha-7 Deficiency','Muscular Dystrophy, Congenital, Type 1D',\n",
    "                                 'Congenital Fibrosis of the Extraocular Muscles']\n",
    "            z_score_list = []\n",
    "            for disgenet_name in disgenet_name_list:\n",
    "                for chem in chem_set:\n",
    "                    try:\n",
    "                        z_score_list.append(exp_proximity_not_significant[chem,disgenet_name][1])\n",
    "                    except:\n",
    "                        pass\n",
    "            if np.mean(z_score_list)>1.5:\n",
    "                country_specific_faraway_dis_dict[dis]=np.mean(z_score_list)\n",
    "            \n",
    "        elif dis=='Congenital birth defects':\n",
    "            disgenet_name_list = [dis for dis in list(diseases_genes_associated_ppi.keys()) if 'Congenital' in dis or 'congenital' in dis]\n",
    "            z_score_list = []\n",
    "            for disgenet_name in disgenet_name_list:\n",
    "                for chem in chem_set:\n",
    "                    try:\n",
    "                        z_score_list.append(exp_proximity_not_significant[chem,disgenet_name][1])\n",
    "                    except:\n",
    "                        pass\n",
    "            if np.mean(z_score_list)>1.5:\n",
    "                country_specific_faraway_dis_dict[dis]=np.mean(z_score_list)\n",
    "\n",
    "        elif dis=='Turner syndrome':\n",
    "            disgenet_name_list = [dis for dis in list(diseases_genes_associated_ppi.keys()) if 'Turner' in dis]\n",
    "            z_score_list = []\n",
    "            for disgenet_name in disgenet_name_list:\n",
    "                for chem in chem_set:\n",
    "                    try:\n",
    "                        z_score_list.append(exp_proximity_not_significant[chem,disgenet_name][1])\n",
    "                    except:\n",
    "                        pass\n",
    "            if np.mean(z_score_list)>1.5:\n",
    "                country_specific_faraway_dis_dict[dis]=np.mean(z_score_list)\n",
    "    low_country_faraway_dis_dict[country] = country_specific_faraway_dis_dict\n",
    "\n",
    "low_country_noprox_dis_dict = {}\n",
    "for country,dis_dict in low_country_selected_chem_id_gbd_mean_dict.items():\n",
    "    n_dis = len(dis_dict)\n",
    "    chem_set = low_country_selected_chem_id_dict[country]\n",
    "    country_noprox_dis_list = []\n",
    "    country_specific_noprox_dis_dict = {}\n",
    "    \n",
    "    for gbd_dis,dislist in no_prox_air_disset_gbd_dict_cleaned_inv.items():\n",
    "        if gbd_dis not in dis_dict.keys():\n",
    "            if gbd_dis not in list(low_country_faraway_dis_dict[country].keys()):\n",
    "                country_noprox_dis_list.append(gbd_dis)\n",
    "    for dis in country_noprox_dis_list:\n",
    "        disgenet_name_list = no_prox_air_disset_gbd_dict_cleaned_inv[dis]\n",
    "        z_score_list = []\n",
    "        for chem in chem_set:\n",
    "            for disgenet_name in disgenet_name_list:\n",
    "                try:\n",
    "                    z_score_list.append(exp_proximity_not_significant[chem,disgenet_name][1])\n",
    "                except:\n",
    "                    pass\n",
    "        if np.mean(z_score_list)>-1.5:\n",
    "            country_specific_noprox_dis_dict[dis]=np.mean(z_score_list)\n",
    "    country_specific_noprox_dis_dict_sort = {k: v for k, v in sorted(country_specific_noprox_dis_dict.items(), key=lambda item: item[1])}\n",
    "    \n",
    "    if len(country_specific_noprox_dis_dict_sort)>n_dis:\n",
    "        country_specific_noprox_dis_dict_sort_keys = list(country_specific_noprox_dis_dict_sort.keys())\n",
    "        n_country_specific_noprox_dis_dict_sort = {}\n",
    "        for k in list(country_specific_noprox_dis_dict_sort_keys)[:n_dis]:\n",
    "            n_country_specific_noprox_dis_dict_sort[k]=country_specific_noprox_dis_dict_sort[k]\n",
    "        \n",
    "        low_country_noprox_dis_dict[country] = n_country_specific_noprox_dis_dict_sort\n",
    "    else:\n",
    "        low_country_noprox_dis_dict[country] = country_specific_noprox_dis_dict_sort\n",
    "\n",
    "#Let's just merge the no-proximal dictionary and the faraway dictionary\n",
    "new_low_country_distant_dis_dict = {}\n",
    "for country,dis_dict in low_country_selected_chem_id_gbd_mean_dict.items():\n",
    "    n_proxdis = len(dis_dict)\n",
    "    n_distantdis = len(low_country_faraway_dis_dict[country])\n",
    "    noprox_list = list(low_country_noprox_dis_dict[country].keys())\n",
    "    missing_dis = n_proxdis-n_distantdis\n",
    "    try:\n",
    "        dislist = noprox_list[:missing_dis]\n",
    "        distant_dict = low_country_faraway_dis_dict[country].copy()\n",
    "        for dis in dislist:\n",
    "            distant_dict[dis] = low_country_noprox_dis_dict[country][dis]\n",
    "        new_low_country_distant_dis_dict[country] = distant_dict\n",
    "    except:\n",
    "        distant_dict = {**low_country_faraway_dis_dict[country], **low_country_noprox_dis_dict[country]}\n",
    "        new_low_country_distant_dis_dict[country] = distant_dict\n",
    "\n",
    "distant_dis_list=[] \n",
    "for country, disdict in new_low_country_distant_dis_dict.items():\n",
    "    distant_dis_list.extend(list(disdict.keys()))\n",
    "    \n",
    "distant_dis_incidence={}\n",
    "country_set=set(GBD_df['location'])\n",
    "for dis in set(distant_dis_list):\n",
    "    incidence_list=[]\n",
    "    for country in country_set:\n",
    "        incidence_list.append(country_dis_incidence[country][dis])\n",
    "    distant_dis_incidence[dis]=incidence_list\n",
    "   \n",
    "distant_dis_prevalence={}\n",
    "for dis in set(distant_dis_list):\n",
    "    prevalence_list=[]\n",
    "    for country in country_set:\n",
    "        prevalence_list.append(country_dis_prevalence[country][dis])\n",
    "    distant_dis_prevalence[dis]=prevalence_list\n",
    "\n",
    "distant_dis_mortality={}\n",
    "for dis in set(distant_dis_list):\n",
    "    mortality_list=[]\n",
    "    for country in country_set:\n",
    "        mortality_list.append(country_dis_mortality[country][dis])\n",
    "    distant_dis_mortality[dis]=mortality_list\n",
    "\n",
    "distant_dis_daly={}\n",
    "for dis in set(distant_dis_list):\n",
    "    daly_list=[]\n",
    "    for country in country_set:\n",
    "        daly_list.append(country_dis_DALYs[country][dis])\n",
    "    distant_dis_daly[dis]=daly_list\n",
    "air_low_country_distant_zscore_dict={}\n",
    "\n",
    "for country,dis_dict in new_low_country_distant_dis_dict.items():\n",
    "    country_distant_zscore_list=[]\n",
    "    for dis in list(dis_dict.keys()):\n",
    "        mu = np.mean(distant_dis_incidence[dis])\n",
    "        std = np.std(distant_dis_incidence[dis])\n",
    "\n",
    "        z = (country_dis_incidence[country][dis]-mu)/std\n",
    "        country_distant_zscore_list.append(z)\n",
    "\n",
    "        from scipy.stats import norm\n",
    "        ptemp = norm.sf(abs(z))\n",
    "    air_low_country_distant_zscore_dict[country]=country_distant_zscore_list"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "4a73dae1-216a-48f5-9375-879b67615f5f",
   "metadata": {},
   "outputs": [],
   "source": [
    "air_pollution_low_zscore_df = pd.DataFrame()\n",
    "\n",
    "selected_countries=set(low_country_proximal_zscore_dict.keys())\n",
    "tot_country_list=[]\n",
    "tot_zscore_list=[]\n",
    "location_network_list=[]\n",
    "for country in selected_countries:\n",
    "    farawaylist_cleaned = [x for x in air_low_country_distant_zscore_dict[country] if str(x) != 'nan']\n",
    "    proximallist_cleaned = [x for x in low_country_proximal_zscore_dict[country] if str(x) != 'nan']\n",
    "    tot_zscorelist_partial=farawaylist_cleaned+proximallist_cleaned\n",
    "    for i in proximallist_cleaned:\n",
    "        location_network_list.append(\"proximal\")\n",
    "        tot_zscore_list.append(i)\n",
    "        tot_country_list.append(country)\n",
    "    for i in farawaylist_cleaned:\n",
    "        location_network_list.append(\"distant\")\n",
    "        tot_zscore_list.append(i)\n",
    "        tot_country_list.append(country)\n",
    "\n",
    "air_pollution_low_zscore_df['location']=tot_country_list\n",
    "air_pollution_low_zscore_df['z-score']=tot_zscore_list\n",
    "air_pollution_low_zscore_df['network_location']=location_network_list\n",
    "\n",
    "my_order = sorted(list(set(air_pollution_low_zscore_df['location'])))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "18b6a69b-8607-403a-9297-223944812085",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Let's test the difference between proximal and no-proximal (because with distant is kind of obvious and it doesn't\n",
    "#make so much sense)\n",
    "import scipy\n",
    "for country in low_country_proximal_zscore_dict.keys():\n",
    "    try:\n",
    "        proximallist_cleaned = [x for x in low_country_proximal_zscore_dict[country] if str(x) != 'nan']\n",
    "        distantlist_cleaned = [x for x in air_low_country_distant_zscore_dict[country] if str(x) != 'nan']\n",
    "        prox_mean = np.mean(proximallist_cleaned)\n",
    "        distant_mean = np.mean(distantlist_cleaned)\n",
    "        pval_noprox = scipy.stats.mannwhitneyu(proximallist_cleaned, distantlist_cleaned, use_continuity=True, alternative='greater')[1]\n",
    "        print(country,prox_mean,distant_mean,round(pval_noprox,4),round(pval_distant,4))\n",
    "    except:\n",
    "        print(country,prox_mean,distant_mean,round(pval_noprox,4))\n",
    "        "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "399240fc-3fa1-426a-a7c3-a6fb94e1b5c0",
   "metadata": {},
   "outputs": [],
   "source": [
    "import scipy \n",
    "# Assuming your data loading and plot setup code here\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(15, 20))\n",
    "sns.set(style=\"white\")\n",
    "g = sns.boxplot(x=\"z-score\", y=\"location\", orient=\"h\",\n",
    "                hue=\"network_location\", palette=[\"#2B7A98\", \"#54CAF2\"], order=my_order,\n",
    "                data=air_pollution_low_zscore_df)\n",
    "fontsize = 36\n",
    "plt.legend(loc=\"upper right\", frameon=True, fontsize=fontsize)\n",
    "plt.yticks(fontsize=36)\n",
    "\n",
    "for ax in plt.gcf().axes:\n",
    "    l = ax.get_xlabel()\n",
    "    ax.set_xlabel(l, fontsize=36)\n",
    "    ax.set_ylabel(l, fontsize=36)\n",
    "\n",
    "plt.ylabel('', fontsize=25)\n",
    "plt.xlabel('', fontsize=25)\n",
    "ax.spines['top'].set_visible(False)\n",
    "ax.spines['right'].set_visible(False)\n",
    "ax.spines['bottom'].set_visible(True)\n",
    "ax.spines['left'].set_visible(True)\n",
    "ax.set_xticklabels(ax.get_xticks(), size=36)\n",
    "\n",
    "# Example of adding significance stars between two boxplots\n",
    "# Let's assume you've got p-values and locations. You'll\n",
    "# Placeholder list of p-values, x (z-score) and y (location) coordinates for the text, and the text to be added (stars).\n",
    "# You need to calculate these based on your actual data and tests.\n",
    "pv_list = []  # Example p-values\n",
    "locations = []  # Example y-axis categories\n",
    "stars = []  # Based on the p-values, this is what we want to add\n",
    "\n",
    "\n",
    "\n",
    "for country in my_order:\n",
    "    proximallist_cleaned = [x for x in low_country_proximal_zscore_dict[country] if str(x) != 'nan']\n",
    "    distantlist_cleaned = [x for x in air_low_country_distant_zscore_dict[country] if str(x) != 'nan']\n",
    "    prox_mean = np.mean(proximallist_cleaned)\n",
    "    distant_mean = np.mean(distantlist_cleaned)\n",
    "    pval_noprox = scipy.stats.mannwhitneyu(proximallist_cleaned, distantlist_cleaned, use_continuity=True, alternative='greater')[1]\n",
    "    pv_list.append(pval_noprox)\n",
    "    locations.append(country)\n",
    "    if pval_noprox<0.001:\n",
    "        stars.append(\"***\")\n",
    "    elif pval_noprox>0.001 and pval_noprox<0.01:\n",
    "        stars.append(\"**\")\n",
    "    elif pval_noprox>0.01 and pval_noprox<0.05:\n",
    "        stars.append(\"*\")\n",
    "    else:\n",
    "        stars.append(\"n.s.\")\n",
    "\n",
    "\n",
    "for pv, loc, star in zip(pv_list, locations, stars):\n",
    "    # This part is tricky without the data; you need to find the specific y coordinates for your boxplots.\n",
    "    # A simplistic approach:\n",
    "    yloc = my_order.index(loc) + 0.4  # This presumes 'my_order' list contains the order of y-categories\n",
    "    xloc = max(air_pollution_low_zscore_df[air_pollution_low_zscore_df['location'] == loc]['z-score']) + 0.5  \n",
    "  \n",
    "    # Add the annotation\n",
    "    ax.text(xloc, yloc, star, fontsize=fontsize, color=\"black\", ha='center')\n",
    "plt.savefig('output/Figures/SF11A_v01_post_rev.pdf',format='pdf',dpi=300, bbox_inches = \"tight\")\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "6cd00564-c8a6-49f1-8d90-cd5cb5e818cd",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Prevalence\n",
    "\n",
    "country_proximal_zscore_dict={}\n",
    "for country,dis_dict in low_country_selected_chem_id_gbd_mean_dict.items():\n",
    "    country_proximal_zscore_list=[]\n",
    "    for dis in list(dis_dict.keys()):\n",
    "        mu = np.mean(proximal_dis_prevalence[dis])\n",
    "        std = np.std(proximal_dis_prevalence[dis])\n",
    "        z = (country_dis_prevalence[country][dis]-mu)/std\n",
    "        country_proximal_zscore_list.append(z)\n",
    "        from scipy.stats import norm\n",
    "        ptemp = norm.sf(abs(z))\n",
    "    if len(country_proximal_zscore_list)>0:\n",
    "        country_proximal_zscore_dict[country]=country_proximal_zscore_list\n",
    "\n",
    "air_country_distant_zscore_dict={}\n",
    "for country,dis_dict in new_low_country_distant_dis_dict.items():\n",
    "    country_distant_zscore_list=[]\n",
    "    for dis in list(dis_dict.keys()):\n",
    "        mu = np.mean(distant_dis_prevalence[dis])\n",
    "        std = np.std(distant_dis_prevalence[dis])\n",
    "        z = (country_dis_prevalence[country][dis]-mu)/std\n",
    "        country_distant_zscore_list.append(z)\n",
    "        from scipy.stats import norm\n",
    "        ptemp = norm.sf(abs(z))\n",
    "    air_country_distant_zscore_dict[country]=country_distant_zscore_list\n",
    "\n",
    "air_pollution_zscore_df = pd.DataFrame()\n",
    "selected_countries=set(country_proximal_zscore_dict.keys())\n",
    "tot_country_list=[]\n",
    "tot_zscore_list=[]\n",
    "location_network_list=[]\n",
    "for country in selected_countries:\n",
    "    farawaylist_cleaned = [x for x in air_country_distant_zscore_dict[country] if str(x) != 'nan']\n",
    "    proximallist_cleaned = [x for x in country_proximal_zscore_dict[country] if str(x) != 'nan']\n",
    "    tot_zscorelist_partial=farawaylist_cleaned+proximallist_cleaned\n",
    "    for i in proximallist_cleaned:\n",
    "        location_network_list.append(\"proximal\")\n",
    "        tot_zscore_list.append(i)\n",
    "        tot_country_list.append(country)\n",
    "    for i in farawaylist_cleaned:\n",
    "        location_network_list.append(\"distant\")\n",
    "        tot_zscore_list.append(i)\n",
    "        tot_country_list.append(country)\n",
    "\n",
    "air_pollution_zscore_df['location']=tot_country_list\n",
    "air_pollution_zscore_df['z-score']=tot_zscore_list\n",
    "air_pollution_zscore_df['network_location']=location_network_list\n",
    "\n",
    "my_order = sorted(list(set(air_pollution_zscore_df['location'])))\n",
    "\n",
    "for country in country_proximal_zscore_dict.keys():\n",
    "    try:\n",
    "        proximallist_cleaned = [x for x in country_proximal_zscore_dict[country] if str(x) != 'nan']\n",
    "        distantlist_cleaned = [x for x in air_country_distant_zscore_dict[country] if str(x) != 'nan']\n",
    "        prox_mean = np.mean(proximallist_cleaned)\n",
    "        distant_mean = np.mean(distantlist_cleaned)\n",
    "        pval_noprox = scipy.stats.mannwhitneyu(proximallist_cleaned, distantlist_cleaned, use_continuity=True, alternative='greater')[1]\n",
    "        print(country,prox_mean,distant_mean,round(pval_noprox,4),round(pval_distant,4))\n",
    "    except:\n",
    "        print(country,prox_mean,distant_mean,round(pval_noprox,4))\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "0973ec2f-765c-4052-915d-a61699851c75",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "fig, ax = plt.subplots(figsize=(15, 20))\n",
    "sns.set(style=\"white\")\n",
    "g = sns.boxplot(x=\"z-score\", y=\"location\", orient=\"h\",\n",
    "                hue=\"network_location\", palette=[\"#2B7A98\", \"#54CAF2\"], order=my_order,\n",
    "                data=air_pollution_zscore_df)\n",
    "fontsize = 36\n",
    "plt.legend(loc=\"upper right\", frameon=True, fontsize=fontsize)\n",
    "plt.yticks(fontsize=36)\n",
    "\n",
    "for ax in plt.gcf().axes:\n",
    "    l = ax.get_xlabel()\n",
    "    ax.set_xlabel(l, fontsize=36)\n",
    "    ax.set_ylabel(l, fontsize=36)\n",
    "\n",
    "plt.ylabel('', fontsize=25)\n",
    "plt.xlabel('', fontsize=25)\n",
    "ax.spines['top'].set_visible(False)\n",
    "ax.spines['right'].set_visible(False)\n",
    "ax.spines['bottom'].set_visible(True)\n",
    "ax.spines['left'].set_visible(True)\n",
    "ax.set_xticklabels(ax.get_xticks(), size=36)\n",
    "\n",
    "pv_list = []  \n",
    "locations = []  \n",
    "stars = []  \n",
    "\n",
    "for country in my_order:\n",
    "    proximallist_cleaned = [x for x in country_proximal_zscore_dict[country] if str(x) != 'nan']\n",
    "    distantlist_cleaned = [x for x in air_country_distant_zscore_dict[country] if str(x) != 'nan']\n",
    "    prox_mean = np.mean(proximallist_cleaned)\n",
    "    distant_mean = np.mean(distantlist_cleaned)\n",
    "    pval_noprox = scipy.stats.mannwhitneyu(proximallist_cleaned, distantlist_cleaned, use_continuity=True, alternative='greater')[1]\n",
    "    pv_list.append(pval_noprox)\n",
    "    locations.append(country)\n",
    "    if pval_noprox<0.001:\n",
    "        stars.append(\"***\")\n",
    "    elif pval_noprox>0.001 and pval_noprox<0.01:\n",
    "        stars.append(\"**\")\n",
    "    elif pval_noprox>0.01 and pval_noprox<0.05:\n",
    "        stars.append(\"*\")\n",
    "    else:\n",
    "        stars.append(\"n.s.\")\n",
    "\n",
    "for pv, loc, star in zip(pv_list, locations, stars):\n",
    "    yloc = my_order.index(loc) + 0.4 \n",
    "    xloc = max(air_pollution_zscore_df[air_pollution_zscore_df['location'] == loc]['z-score']) + 0.5  \n",
    "    ax.text(xloc, yloc, star, fontsize=fontsize, color=\"black\", ha='center')\n",
    "\n",
    "plt.savefig('output/Figures/SF12B_v01_post_rev.pdf',format='pdf',dpi=300, bbox_inches = \"tight\")\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "1b3d3224-53ae-4dd2-9817-aeebbfe3a4f4",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "04ae9606-e785-47dc-86cc-d9398236fb64",
   "metadata": {},
   "outputs": [],
   "source": [
    "#mortality\n",
    "\n",
    "country_proximal_zscore_dict={}\n",
    "for country,dis_dict in low_country_selected_chem_id_gbd_mean_dict.items():\n",
    "    country_proximal_zscore_list=[]\n",
    "    for dis in list(dis_dict.keys()):\n",
    "        mu = np.mean(proximal_dis_mortality[dis])\n",
    "        std = np.std(proximal_dis_mortality[dis])\n",
    "        z = (country_dis_mortality[country][dis]-mu)/std\n",
    "        country_proximal_zscore_list.append(z)\n",
    "        from scipy.stats import norm\n",
    "        ptemp = norm.sf(abs(z))\n",
    "    if len(country_proximal_zscore_list)>0:\n",
    "        country_proximal_zscore_dict[country]=country_proximal_zscore_list\n",
    "\n",
    "air_country_distant_zscore_dict={}\n",
    "for country,dis_dict in new_low_country_distant_dis_dict.items():\n",
    "    country_distant_zscore_list=[]\n",
    "    for dis in list(dis_dict.keys()):\n",
    "        mu = np.mean(distant_dis_mortality[dis])\n",
    "        std = np.std(distant_dis_mortality[dis])\n",
    "        z = (country_dis_mortality[country][dis]-mu)/std\n",
    "        country_distant_zscore_list.append(z)\n",
    "        from scipy.stats import norm\n",
    "        ptemp = norm.sf(abs(z))\n",
    "    air_country_distant_zscore_dict[country]=country_distant_zscore_list\n",
    "\n",
    "air_pollution_zscore_df = pd.DataFrame()\n",
    "selected_countries=set(country_proximal_zscore_dict.keys())\n",
    "tot_country_list=[]\n",
    "tot_zscore_list=[]\n",
    "location_network_list=[]\n",
    "for country in selected_countries:\n",
    "    farawaylist_cleaned = [x for x in air_country_distant_zscore_dict[country] if str(x) != 'nan']\n",
    "    proximallist_cleaned = [x for x in country_proximal_zscore_dict[country] if str(x) != 'nan']\n",
    "    tot_zscorelist_partial=farawaylist_cleaned+proximallist_cleaned\n",
    "    for i in proximallist_cleaned:\n",
    "        location_network_list.append(\"proximal\")\n",
    "        tot_zscore_list.append(i)\n",
    "        tot_country_list.append(country)\n",
    "    for i in farawaylist_cleaned:\n",
    "        location_network_list.append(\"distant\")\n",
    "        tot_zscore_list.append(i)\n",
    "        tot_country_list.append(country)\n",
    "\n",
    "air_pollution_zscore_df['location']=tot_country_list\n",
    "air_pollution_zscore_df['z-score']=tot_zscore_list\n",
    "air_pollution_zscore_df['network_location']=location_network_list\n",
    "\n",
    "my_order = sorted(list(set(air_pollution_zscore_df['location'])))\n",
    "\n",
    "for country in country_proximal_zscore_dict.keys():\n",
    "    try:\n",
    "        proximallist_cleaned = [x for x in country_proximal_zscore_dict[country] if str(x) != 'nan']\n",
    "        distantlist_cleaned = [x for x in air_country_distant_zscore_dict[country] if str(x) != 'nan']\n",
    "        prox_mean = np.mean(proximallist_cleaned)\n",
    "        distant_mean = np.mean(distantlist_cleaned)\n",
    "        pval_noprox = scipy.stats.mannwhitneyu(proximallist_cleaned, distantlist_cleaned, use_continuity=True, alternative='greater')[1]\n",
    "        print(country,prox_mean,distant_mean,round(pval_noprox,4),round(pval_distant,4))\n",
    "    except:\n",
    "        print(country,prox_mean,distant_mean,round(pval_noprox,4))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "86fab565-b792-4260-8bef-ef8345bab75f",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "fig, ax = plt.subplots(figsize=(15, 20))\n",
    "sns.set(style=\"white\")\n",
    "g = sns.boxplot(x=\"z-score\", y=\"location\", orient=\"h\",\n",
    "                hue=\"network_location\", palette=[\"#2B7A98\", \"#54CAF2\"], order=my_order,\n",
    "                data=air_pollution_zscore_df)\n",
    "fontsize = 36\n",
    "plt.legend(loc=\"upper right\", frameon=True, fontsize=fontsize)\n",
    "plt.yticks(fontsize=36)\n",
    "\n",
    "for ax in plt.gcf().axes:\n",
    "    l = ax.get_xlabel()\n",
    "    ax.set_xlabel(l, fontsize=36)\n",
    "    ax.set_ylabel(l, fontsize=36)\n",
    "\n",
    "plt.ylabel('', fontsize=25)\n",
    "plt.xlabel('', fontsize=25)\n",
    "ax.spines['top'].set_visible(False)\n",
    "ax.spines['right'].set_visible(False)\n",
    "ax.spines['bottom'].set_visible(True)\n",
    "ax.spines['left'].set_visible(True)\n",
    "ax.set_xticklabels(ax.get_xticks(), size=36)\n",
    "\n",
    "pv_list = []  \n",
    "locations = []  \n",
    "stars = []  \n",
    "\n",
    "for country in my_order:\n",
    "    proximallist_cleaned = [x for x in country_proximal_zscore_dict[country] if str(x) != 'nan']\n",
    "    distantlist_cleaned = [x for x in air_country_distant_zscore_dict[country] if str(x) != 'nan']\n",
    "    prox_mean = np.mean(proximallist_cleaned)\n",
    "    distant_mean = np.mean(distantlist_cleaned)\n",
    "    pval_noprox = scipy.stats.mannwhitneyu(proximallist_cleaned, distantlist_cleaned, use_continuity=True, alternative='greater')[1]\n",
    "    pv_list.append(pval_noprox)\n",
    "    locations.append(country)\n",
    "    if pval_noprox<0.001:\n",
    "        stars.append(\"***\")\n",
    "    elif pval_noprox>0.001 and pval_noprox<0.01:\n",
    "        stars.append(\"**\")\n",
    "    elif pval_noprox>0.01 and pval_noprox<0.05:\n",
    "        stars.append(\"*\")\n",
    "    else:\n",
    "        stars.append(\"n.s.\")\n",
    "\n",
    "for pv, loc, star in zip(pv_list, locations, stars):\n",
    "    yloc = my_order.index(loc) + 0.4 \n",
    "    xloc = max(air_pollution_zscore_df[air_pollution_zscore_df['location'] == loc]['z-score']) + 0.5  \n",
    "    ax.text(xloc, yloc, star, fontsize=fontsize, color=\"black\", ha='center')\n",
    "\n",
    "plt.savefig('output/Figures/SF13B_v01_post_rev.pdf',format='pdf',dpi=300, bbox_inches = \"tight\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "95c237ae-7da9-4ed7-ab5a-8b01188c236b",
   "metadata": {},
   "outputs": [],
   "source": [
    "#daly\n",
    "\n",
    "country_proximal_zscore_dict={}\n",
    "for country,dis_dict in low_country_selected_chem_id_gbd_mean_dict.items():\n",
    "    country_proximal_zscore_list=[]\n",
    "    for dis in list(dis_dict.keys()):\n",
    "        mu = np.mean(proximal_dis_daly[dis])\n",
    "        std = np.std(proximal_dis_daly[dis])\n",
    "        z = (country_dis_DALYs[country][dis]-mu)/std\n",
    "        country_proximal_zscore_list.append(z)\n",
    "        from scipy.stats import norm\n",
    "        ptemp = norm.sf(abs(z))\n",
    "    if len(country_proximal_zscore_list)>0:\n",
    "        country_proximal_zscore_dict[country]=country_proximal_zscore_list\n",
    "\n",
    "air_country_distant_zscore_dict={}\n",
    "for country,dis_dict in new_low_country_distant_dis_dict.items():\n",
    "    country_distant_zscore_list=[]\n",
    "    for dis in list(dis_dict.keys()):\n",
    "        mu = np.mean(distant_dis_daly[dis])\n",
    "        std = np.std(distant_dis_daly[dis])\n",
    "        z = (country_dis_DALYs[country][dis]-mu)/std\n",
    "        country_distant_zscore_list.append(z)\n",
    "        from scipy.stats import norm\n",
    "        ptemp = norm.sf(abs(z))\n",
    "    air_country_distant_zscore_dict[country]=country_distant_zscore_list\n",
    "\n",
    "air_pollution_zscore_df = pd.DataFrame()\n",
    "selected_countries=set(country_proximal_zscore_dict.keys())\n",
    "tot_country_list=[]\n",
    "tot_zscore_list=[]\n",
    "location_network_list=[]\n",
    "for country in selected_countries:\n",
    "    farawaylist_cleaned = [x for x in air_country_distant_zscore_dict[country] if str(x) != 'nan']\n",
    "    proximallist_cleaned = [x for x in country_proximal_zscore_dict[country] if str(x) != 'nan']\n",
    "    tot_zscorelist_partial=farawaylist_cleaned+proximallist_cleaned\n",
    "    for i in proximallist_cleaned:\n",
    "        location_network_list.append(\"proximal\")\n",
    "        tot_zscore_list.append(i)\n",
    "        tot_country_list.append(country)\n",
    "    for i in farawaylist_cleaned:\n",
    "        location_network_list.append(\"distant\")\n",
    "        tot_zscore_list.append(i)\n",
    "        tot_country_list.append(country)\n",
    "\n",
    "air_pollution_zscore_df['location']=tot_country_list\n",
    "air_pollution_zscore_df['z-score']=tot_zscore_list\n",
    "air_pollution_zscore_df['network_location']=location_network_list\n",
    "\n",
    "my_order = sorted(list(set(air_pollution_zscore_df['location'])))\n",
    "\n",
    "for country in country_proximal_zscore_dict.keys():\n",
    "    try:\n",
    "        proximallist_cleaned = [x for x in country_proximal_zscore_dict[country] if str(x) != 'nan']\n",
    "        distantlist_cleaned = [x for x in air_country_distant_zscore_dict[country] if str(x) != 'nan']\n",
    "        prox_mean = np.mean(proximallist_cleaned)\n",
    "        distant_mean = np.mean(distantlist_cleaned)\n",
    "        pval_noprox = scipy.stats.mannwhitneyu(proximallist_cleaned, distantlist_cleaned, use_continuity=True, alternative='greater')[1]\n",
    "        print(country,prox_mean,distant_mean,round(pval_noprox,4),round(pval_distant,4))\n",
    "    except:\n",
    "        print(country,prox_mean,distant_mean,round(pval_noprox,4))\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "26c3ee88-995a-4a21-ad7f-1b9d2b57541b",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "fig, ax = plt.subplots(figsize=(15, 20))\n",
    "sns.set(style=\"white\")\n",
    "g = sns.boxplot(x=\"z-score\", y=\"location\", orient=\"h\",\n",
    "                hue=\"network_location\", palette=[\"#2B7A98\", \"#54CAF2\"], order=my_order,\n",
    "                data=air_pollution_zscore_df)\n",
    "fontsize = 36\n",
    "plt.legend(loc=\"upper right\", frameon=True, fontsize=fontsize)\n",
    "plt.yticks(fontsize=36)\n",
    "\n",
    "for ax in plt.gcf().axes:\n",
    "    l = ax.get_xlabel()\n",
    "    ax.set_xlabel(l, fontsize=36)\n",
    "    ax.set_ylabel(l, fontsize=36)\n",
    "\n",
    "plt.ylabel('', fontsize=25)\n",
    "plt.xlabel('', fontsize=25)\n",
    "ax.spines['top'].set_visible(False)\n",
    "ax.spines['right'].set_visible(False)\n",
    "ax.spines['bottom'].set_visible(True)\n",
    "ax.spines['left'].set_visible(True)\n",
    "ax.set_xticklabels(ax.get_xticks(), size=36)\n",
    "\n",
    "pv_list = []  \n",
    "locations = []  \n",
    "stars = []  \n",
    "\n",
    "for country in my_order:\n",
    "    proximallist_cleaned = [x for x in country_proximal_zscore_dict[country] if str(x) != 'nan']\n",
    "    distantlist_cleaned = [x for x in air_country_distant_zscore_dict[country] if str(x) != 'nan']\n",
    "    prox_mean = np.mean(proximallist_cleaned)\n",
    "    distant_mean = np.mean(distantlist_cleaned)\n",
    "    pval_noprox = scipy.stats.mannwhitneyu(proximallist_cleaned, distantlist_cleaned, use_continuity=True, alternative='greater')[1]\n",
    "    pv_list.append(pval_noprox)\n",
    "    locations.append(country)\n",
    "    if pval_noprox<0.001:\n",
    "        stars.append(\"***\")\n",
    "    elif pval_noprox>0.001 and pval_noprox<0.01:\n",
    "        stars.append(\"**\")\n",
    "    elif pval_noprox>0.01 and pval_noprox<0.05:\n",
    "        stars.append(\"*\")\n",
    "    else:\n",
    "        stars.append(\"n.s.\")\n",
    "\n",
    "for pv, loc, star in zip(pv_list, locations, stars):\n",
    "    yloc = my_order.index(loc) + 0.4 \n",
    "    xloc = max(air_pollution_zscore_df[air_pollution_zscore_df['location'] == loc]['z-score']) + 0.5  \n",
    "    ax.text(xloc, yloc, star, fontsize=fontsize, color=\"black\", ha='center')\n",
    "\n",
    "plt.savefig('output/Figures/SF14B_v01_post_rev.pdf',format='pdf',dpi=300, bbox_inches = \"tight\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "85fd5dd8-7cd8-4e06-b016-918c66498fdc",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "72cd94b4-1038-491b-bbd0-8746f947d967",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "025a4410-b859-4d50-bb62-bb9c1926c8ec",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "12a9bb7c-005a-485e-8349-b966719fd543",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "6155b312-2feb-4dcb-88be-aa080562ca78",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "cedd88ec-5d1e-4414-be4f-cf9c258f3111",
   "metadata": {},
   "outputs": [],
   "source": [
    "#HIGH\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d2cf6fda-71c8-45e7-8b0c-d5dd0f71b92b",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Proximal\n",
    "high_country_selected_dis_mean_prox_dict = {}\n",
    "for country,disset in high_country_selected_dis_dict.items():\n",
    "    dis_dict = {}\n",
    "    for dis in disset:\n",
    "        z_list = [] \n",
    "        if len(high_country_selected_chem_id_dict[country])>0:\n",
    "            for chem in high_country_selected_chem_id_dict[country]:\n",
    "                try:\n",
    "                    z_list.append(real_significant_final_exp_proximity[chem,dis][1])\n",
    "                except:\n",
    "                    pass\n",
    "        dis_dict[dis]= np.round(np.mean(z_list),2)\n",
    "                    \n",
    "    high_country_selected_dis_mean_prox_dict[country] = dis_dict\n",
    "\n",
    "high_country_selected_chem_id_gbd_dict = {}\n",
    "for country,dis_dict in high_country_selected_dis_mean_prox_dict.items():\n",
    "    if len(dis_dict)>0:\n",
    "        dis_z_list = []\n",
    "        for dis,z in dis_dict.items():\n",
    "            try:\n",
    "                dis_z_list.append([air_disgenet_gbd_dict[dis],z])\n",
    "            except:\n",
    "                pass\n",
    "        high_country_selected_chem_id_gbd_dict[country] = dis_z_list\n",
    "\n",
    "high_country_selected_chem_id_gbd_mean_dict = {}\n",
    "for country,dislist in high_country_selected_chem_id_gbd_dict.items():\n",
    "    dis_dict = {}\n",
    "    dis_list = []\n",
    "    for el in dislist:\n",
    "        dis_list.append(el[0])\n",
    "    dis_list_counter_dict = dict(Counter(dis_list).most_common())\n",
    "    for dis,count in dis_list_counter_dict.items():\n",
    "        z_list = []\n",
    "        if count>1:\n",
    "            for el in dislist:\n",
    "                if dis==el[0]:\n",
    "                    z_list.append(el[1])\n",
    "            dis_dict[dis]=np.round(np.mean(z_list),2)\n",
    "        else:\n",
    "            for el in dislist:\n",
    "                if dis==el[0]:\n",
    "                    dis_dict[el[0]] = el[1]\n",
    "    high_country_selected_chem_id_gbd_mean_dict[country]=dis_dict\n",
    "\n",
    "proximal_dis_list=[] \n",
    "for country, disdict in high_country_selected_chem_id_gbd_mean_dict.items():\n",
    "    proximal_dis_list.extend(list(disdict.keys()))\n",
    "    \n",
    "proximal_dis_incidence={}\n",
    "country_set=set(GBD_df['location'])\n",
    "for dis in set(proximal_dis_list):\n",
    "    incidence_list=[]\n",
    "    for country in country_set:\n",
    "        incidence_list.append(country_dis_incidence[country][dis])\n",
    "    proximal_dis_incidence[dis]=incidence_list\n",
    "\n",
    "proximal_dis_prevalence={}\n",
    "for dis in set(proximal_dis_list):\n",
    "    prevalence_list=[]\n",
    "    for country in country_set:\n",
    "        prevalence_list.append(country_dis_prevalence[country][dis])\n",
    "    proximal_dis_prevalence[dis]=prevalence_list\n",
    "\n",
    "proximal_dis_mortality={}\n",
    "for dis in set(proximal_dis_list):\n",
    "    mortality_list=[]\n",
    "    for country in country_set:\n",
    "        mortality_list.append(country_dis_mortality[country][dis])\n",
    "    proximal_dis_mortality[dis]=mortality_list\n",
    "\n",
    "proximal_dis_daly={}\n",
    "for dis in set(proximal_dis_list):\n",
    "    daly_list=[]\n",
    "    for country in country_set:\n",
    "        daly_list.append(country_dis_DALYs[country][dis])\n",
    "    proximal_dis_daly[dis]=daly_list\n",
    "\n",
    "\n",
    "high_country_proximal_zscore_dict={}\n",
    "for country,dis_dict in high_country_selected_chem_id_gbd_mean_dict.items():\n",
    "    country_proximal_zscore_list=[]\n",
    "    for dis in list(dis_dict.keys()):\n",
    "        mu = np.mean(proximal_dis_incidence[dis])\n",
    "        std = np.std(proximal_dis_incidence[dis])\n",
    "\n",
    "        z = (country_dis_incidence[country][dis]-mu)/std\n",
    "        country_proximal_zscore_list.append(z)\n",
    "\n",
    "        from scipy.stats import norm\n",
    "        ptemp = norm.sf(abs(z))\n",
    "    if len(country_proximal_zscore_list)>0:\n",
    "        high_country_proximal_zscore_dict[country]=country_proximal_zscore_list\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ba33b437-b2a4-4402-94e6-23fae6462f88",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Distant\n",
    "high_country_selected_faraway_dis_dict={}\n",
    "for country,chemset in high_country_selected_chem_id_dict.items():\n",
    "    if len(chemset)>0:\n",
    "        far_away_dis_set = []\n",
    "        for chem in chemset:\n",
    "            for chem_dis_pair in faraway_exp_dis_dict.keys():\n",
    "                if chem in chem_dis_pair[0]:\n",
    "                    far_away_dis_set.append(chem_dis_pair[1])\n",
    "        high_country_selected_faraway_dis_dict[country] = set(far_away_dis_set)\n",
    "\n",
    "\n",
    "high_country_faraway_dis_dict = {}\n",
    "for country,dis_dict in high_country_selected_chem_id_gbd_mean_dict.items():\n",
    "    chem_set = high_country_selected_chem_id_dict[country]\n",
    "    country_faraway_dis_list = []\n",
    "    country_specific_faraway_dis_dict = {}\n",
    "    \n",
    "    for dis in faraway_dis_list:\n",
    "        if dis not in dis_dict.keys():\n",
    "            country_faraway_dis_list.append(dis)\n",
    "    for dis in country_faraway_dis_list:\n",
    "        if dis not in faraway_manually_added_diseases:\n",
    "            disgenet_name_list = []\n",
    "            z_score_list = []\n",
    "            for disgenet_name,gbd_id in n_long_faraway_air_disset_gbd_dict_cleaned.items():\n",
    "                if dis == gbd_id:\n",
    "                    disgenet_name_list.append(disgenet_name)\n",
    "                    for chem in chem_set:\n",
    "                        try:\n",
    "                            z_score_list.append(exp_proximity_not_significant[chem,disgenet_name][1])\n",
    "                        except:\n",
    "                            pass\n",
    "            if np.mean(z_score_list)>1.5:\n",
    "                country_specific_faraway_dis_dict[dis]=np.mean(z_score_list)\n",
    "        elif dis=='Congenital heart anomalies':\n",
    "            disgenet_name_list = ['Congenital Heart Defects','Deafness, Congenital Heart Defects, and Posterior Embryotoxon',\n",
    "                                 'Congenital long QT syndrome']\n",
    "            z_score_list = []\n",
    "            for disgenet_name in disgenet_name_list:\n",
    "                for chem in chem_set:\n",
    "                    try:\n",
    "                        z_score_list.append(exp_proximity_not_significant[chem,disgenet_name][1])\n",
    "                    except:\n",
    "                        pass\n",
    "            if np.mean(z_score_list)>1.5:\n",
    "                country_specific_faraway_dis_dict[dis]=np.mean(z_score_list)\n",
    "        \n",
    "        elif dis=='Congenital musculoskeletal and limb anomalies':\n",
    "            disgenet_name_list = ['Congenital Fiber Type Disproportion','Myopathy, Actin, Congenital, with Excess of Thin Myofilaments',\n",
    "                                 'Myopathy, Actin, Congenital, With Cores','Muscular Dystrophy, Congenital, Megaconial Type',\n",
    "                                 'Myotonia Congenita','Myasthenic Syndromes, Congenital','Muscular Dystrophy, Congenital, Megaconial Type',\n",
    "                                 'Congenital Myasthenic Syndromes, Postsynaptic','Congenital Myasthenic Syndromes, Presynaptic',\n",
    "                                 'Muscular Dystrophy, Congenital, Due To Integrin Alpha-7 Deficiency','Muscular Dystrophy, Congenital, Type 1D',\n",
    "                                 'Congenital Fibrosis of the Extraocular Muscles']\n",
    "            z_score_list = []\n",
    "            for disgenet_name in disgenet_name_list:\n",
    "                for chem in chem_set:\n",
    "                    try:\n",
    "                        z_score_list.append(exp_proximity_not_significant[chem,disgenet_name][1])\n",
    "                    except:\n",
    "                        pass\n",
    "            if np.mean(z_score_list)>1.5:\n",
    "                country_specific_faraway_dis_dict[dis]=np.mean(z_score_list)\n",
    "            \n",
    "        elif dis=='Congenital birth defects':\n",
    "            disgenet_name_list = [dis for dis in list(diseases_genes_associated_ppi.keys()) if 'Congenital' in dis or 'congenital' in dis]\n",
    "            z_score_list = []\n",
    "            for disgenet_name in disgenet_name_list:\n",
    "                for chem in chem_set:\n",
    "                    try:\n",
    "                        z_score_list.append(exp_proximity_not_significant[chem,disgenet_name][1])\n",
    "                    except:\n",
    "                        pass\n",
    "            if np.mean(z_score_list)>1.5:\n",
    "                country_specific_faraway_dis_dict[dis]=np.mean(z_score_list)\n",
    "\n",
    "        elif dis=='Turner syndrome':\n",
    "            disgenet_name_list = [dis for dis in list(diseases_genes_associated_ppi.keys()) if 'Turner' in dis]\n",
    "            z_score_list = []\n",
    "            for disgenet_name in disgenet_name_list:\n",
    "                for chem in chem_set:\n",
    "                    try:\n",
    "                        z_score_list.append(exp_proximity_not_significant[chem,disgenet_name][1])\n",
    "                    except:\n",
    "                        pass\n",
    "            if np.mean(z_score_list)>1.5:\n",
    "                country_specific_faraway_dis_dict[dis]=np.mean(z_score_list)\n",
    "    high_country_faraway_dis_dict[country] = country_specific_faraway_dis_dict\n",
    "\n",
    "high_country_noprox_dis_dict = {}\n",
    "for country,dis_dict in high_country_selected_chem_id_gbd_mean_dict.items():\n",
    "    n_dis = len(dis_dict)\n",
    "    chem_set = high_country_selected_chem_id_dict[country]\n",
    "    country_noprox_dis_list = []\n",
    "    country_specific_noprox_dis_dict = {}\n",
    "    \n",
    "    for gbd_dis,dislist in no_prox_air_disset_gbd_dict_cleaned_inv.items():\n",
    "        if gbd_dis not in dis_dict.keys():\n",
    "            if gbd_dis not in list(high_country_faraway_dis_dict[country].keys()):\n",
    "                country_noprox_dis_list.append(gbd_dis)\n",
    "    for dis in country_noprox_dis_list:\n",
    "        disgenet_name_list = no_prox_air_disset_gbd_dict_cleaned_inv[dis]\n",
    "        z_score_list = []\n",
    "        for chem in chem_set:\n",
    "            for disgenet_name in disgenet_name_list:\n",
    "                try:\n",
    "                    z_score_list.append(exp_proximity_not_significant[chem,disgenet_name][1])\n",
    "                except:\n",
    "                    pass\n",
    "        if np.mean(z_score_list)>-1.5:\n",
    "            country_specific_noprox_dis_dict[dis]=np.mean(z_score_list)\n",
    "    country_specific_noprox_dis_dict_sort = {k: v for k, v in sorted(country_specific_noprox_dis_dict.items(), key=lambda item: item[1])}\n",
    "    \n",
    "    if len(country_specific_noprox_dis_dict_sort)>n_dis:\n",
    "        country_specific_noprox_dis_dict_sort_keys = list(country_specific_noprox_dis_dict_sort.keys())\n",
    "        n_country_specific_noprox_dis_dict_sort = {}\n",
    "        for k in list(country_specific_noprox_dis_dict_sort_keys)[:n_dis]:\n",
    "            n_country_specific_noprox_dis_dict_sort[k]=country_specific_noprox_dis_dict_sort[k]\n",
    "        \n",
    "        high_country_noprox_dis_dict[country] = n_country_specific_noprox_dis_dict_sort\n",
    "    else:\n",
    "        high_country_noprox_dis_dict[country] = country_specific_noprox_dis_dict_sort\n",
    "\n",
    "#Let's just merge the no-proximal dictionary and the faraway dictionary\n",
    "new_high_country_distant_dis_dict = {}\n",
    "for country,dis_dict in high_country_selected_chem_id_gbd_mean_dict.items():\n",
    "    n_proxdis = len(dis_dict)\n",
    "    n_distantdis = len(high_country_faraway_dis_dict[country])\n",
    "    noprox_list = list(high_country_noprox_dis_dict[country].keys())\n",
    "    missing_dis = n_proxdis-n_distantdis\n",
    "    try:\n",
    "        dislist = noprox_list[:missing_dis]\n",
    "        distant_dict = high_country_faraway_dis_dict[country].copy()\n",
    "        for dis in dislist:\n",
    "            distant_dict[dis] = high_country_noprox_dis_dict[country][dis]\n",
    "        new_high_country_distant_dis_dict[country] = distant_dict\n",
    "    except:\n",
    "        distant_dict = {**high_country_faraway_dis_dict[country], **high_country_noprox_dis_dict[country]}\n",
    "        new_high_country_distant_dis_dict[country] = distant_dict\n",
    "\n",
    "distant_dis_list=[] \n",
    "for country, disdict in new_high_country_distant_dis_dict.items():\n",
    "    distant_dis_list.extend(list(disdict.keys()))\n",
    "    \n",
    "distant_dis_incidence={}\n",
    "country_set=set(GBD_df['location'])\n",
    "for dis in set(distant_dis_list):\n",
    "    incidence_list=[]\n",
    "    for country in country_set:\n",
    "        incidence_list.append(country_dis_incidence[country][dis])\n",
    "    distant_dis_incidence[dis]=incidence_list\n",
    "  \n",
    "distant_dis_prevalence={}\n",
    "for dis in set(distant_dis_list):\n",
    "     prevalence_list=[]\n",
    "     for country in country_set:\n",
    "         prevalence_list.append(country_dis_prevalence[country][dis])\n",
    "     distant_dis_prevalence[dis]=prevalence_list\n",
    "\n",
    "distant_dis_mortality={}\n",
    "for dis in set(distant_dis_list):\n",
    "     mortality_list=[]\n",
    "     for country in country_set:\n",
    "         mortality_list.append(country_dis_mortality[country][dis])\n",
    "     distant_dis_mortality[dis]=mortality_list\n",
    "\n",
    "distant_dis_daly={}\n",
    "for dis in set(distant_dis_list):\n",
    "     daly_list=[]\n",
    "     for country in country_set:\n",
    "         daly_list.append(country_dis_DALYs[country][dis])\n",
    "     distant_dis_daly[dis]=daly_list\n",
    "\n",
    "air_high_country_distant_zscore_dict={}\n",
    "\n",
    "for country,dis_dict in new_high_country_distant_dis_dict.items():\n",
    "    country_distant_zscore_list=[]\n",
    "    for dis in list(dis_dict.keys()):\n",
    "        mu = np.mean(distant_dis_incidence[dis])\n",
    "        std = np.std(distant_dis_incidence[dis])\n",
    "\n",
    "        z = (country_dis_incidence[country][dis]-mu)/std\n",
    "        country_distant_zscore_list.append(z)\n",
    "\n",
    "        from scipy.stats import norm\n",
    "        ptemp = norm.sf(abs(z))\n",
    "    air_high_country_distant_zscore_dict[country]=country_distant_zscore_list"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "637eb76b-a3da-4a20-bc54-7c1d242f3794",
   "metadata": {},
   "outputs": [],
   "source": [
    "air_pollution_high_zscore_df = pd.DataFrame()\n",
    "\n",
    "selected_countries=set(high_country_proximal_zscore_dict.keys())\n",
    "tot_country_list=[]\n",
    "tot_zscore_list=[]\n",
    "location_network_list=[]\n",
    "for country in selected_countries:\n",
    "    farawaylist_cleaned = [x for x in air_high_country_distant_zscore_dict[country] if str(x) != 'nan']\n",
    "    proximallist_cleaned = [x for x in high_country_proximal_zscore_dict[country] if str(x) != 'nan']\n",
    "    tot_zscorelist_partial=farawaylist_cleaned+proximallist_cleaned\n",
    "    for i in proximallist_cleaned:\n",
    "        location_network_list.append(\"proximal\")\n",
    "        tot_zscore_list.append(i)\n",
    "        tot_country_list.append(country)\n",
    "    for i in farawaylist_cleaned:\n",
    "        location_network_list.append(\"distant\")\n",
    "        tot_zscore_list.append(i)\n",
    "        tot_country_list.append(country)\n",
    "\n",
    "air_pollution_high_zscore_df['location']=tot_country_list\n",
    "air_pollution_high_zscore_df['z-score']=tot_zscore_list\n",
    "air_pollution_high_zscore_df['network_location']=location_network_list\n",
    "\n",
    "my_order = sorted(list(set(air_pollution_high_zscore_df['location'])))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "adca0093-90dd-4b91-8668-4530297ebfc2",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Let's test the difference between proximal and no-proximal (because with distant is kind of obvious and it doesn't\n",
    "#make so much sense)\n",
    "import scipy\n",
    "for country in high_country_proximal_zscore_dict.keys():\n",
    "    try:\n",
    "        proximallist_cleaned = [x for x in high_country_proximal_zscore_dict[country] if str(x) != 'nan']\n",
    "        distantlist_cleaned = [x for x in air_high_country_distant_zscore_dict[country] if str(x) != 'nan']\n",
    "        prox_mean = np.mean(proximallist_cleaned)\n",
    "        distant_mean = np.mean(distantlist_cleaned)\n",
    "        pval_noprox = scipy.stats.mannwhitneyu(proximallist_cleaned, distantlist_cleaned, use_continuity=True, alternative='greater')[1]\n",
    "        print(country,prox_mean,distant_mean,round(pval_noprox,4),round(pval_distant,4))\n",
    "    except:\n",
    "        print(country,prox_mean,distant_mean,round(pval_noprox,4))\n",
    "        "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "e0fc3fbe-c76c-47e1-aed3-52437b106adf",
   "metadata": {},
   "outputs": [],
   "source": [
    "import scipy \n",
    "# Assuming your data loading and plot setup code here\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(15, 20))\n",
    "sns.set(style=\"white\")\n",
    "g = sns.boxplot(x=\"z-score\", y=\"location\", orient=\"h\",\n",
    "                hue=\"network_location\", palette=[\"#2B7A98\", \"#54CAF2\"], order=my_order,\n",
    "                data=air_pollution_high_zscore_df)\n",
    "fontsize = 36\n",
    "plt.legend(loc=\"upper right\", frameon=True, fontsize=fontsize)\n",
    "plt.yticks(fontsize=36)\n",
    "\n",
    "for ax in plt.gcf().axes:\n",
    "    l = ax.get_xlabel()\n",
    "    ax.set_xlabel(l, fontsize=36)\n",
    "    ax.set_ylabel(l, fontsize=36)\n",
    "\n",
    "plt.ylabel('', fontsize=25)\n",
    "plt.xlabel('', fontsize=25)\n",
    "ax.spines['top'].set_visible(False)\n",
    "ax.spines['right'].set_visible(False)\n",
    "ax.spines['bottom'].set_visible(True)\n",
    "ax.spines['left'].set_visible(True)\n",
    "ax.set_xticklabels(ax.get_xticks(), size=36)\n",
    "\n",
    "pv_list = []  \n",
    "locations = []  \n",
    "stars = []  \n",
    "for country in my_order:\n",
    "    proximallist_cleaned = [x for x in high_country_proximal_zscore_dict[country] if str(x) != 'nan']\n",
    "    distantlist_cleaned = [x for x in air_high_country_distant_zscore_dict[country] if str(x) != 'nan']\n",
    "    prox_mean = np.mean(proximallist_cleaned)\n",
    "    distant_mean = np.mean(distantlist_cleaned)\n",
    "    pval_noprox = scipy.stats.mannwhitneyu(proximallist_cleaned, distantlist_cleaned, use_continuity=True, alternative='greater')[1]\n",
    "    pv_list.append(pval_noprox)\n",
    "    locations.append(country)\n",
    "    if pval_noprox<0.001:\n",
    "        stars.append(\"***\")\n",
    "    elif pval_noprox>0.001 and pval_noprox<0.01:\n",
    "        stars.append(\"**\")\n",
    "    elif pval_noprox>0.01 and pval_noprox<0.05:\n",
    "        stars.append(\"*\")\n",
    "    else:\n",
    "        stars.append(\"n.s.\")\n",
    "\n",
    "\n",
    "for pv, loc, star in zip(pv_list, locations, stars):\n",
    "    yloc = my_order.index(loc) + 0.4 \n",
    "    xloc = max(air_pollution_high_zscore_df[air_pollution_high_zscore_df['location'] == loc]['z-score']) + 0.5  \n",
    "    # Add the annotation\n",
    "    ax.text(xloc, yloc, star, fontsize=fontsize, color=\"black\", ha='center')\n",
    "    \n",
    "plt.savefig('output/Figures/SF11B_v01_post_rev.pdf',format='pdf',dpi=300, bbox_inches = \"tight\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d4fd107b-dc08-4d0b-9b46-7e356c14e947",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "dbbb3644-b2f7-4ad4-abeb-12cf5698aa92",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ce2c2506-7318-45bc-af0d-4ba44f6f81a7",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Prevalence\n",
    "\n",
    "country_proximal_zscore_dict={}\n",
    "for country,dis_dict in high_country_selected_chem_id_gbd_mean_dict.items():\n",
    "    country_proximal_zscore_list=[]\n",
    "    for dis in list(dis_dict.keys()):\n",
    "        mu = np.mean(proximal_dis_prevalence[dis])\n",
    "        std = np.std(proximal_dis_prevalence[dis])\n",
    "        z = (country_dis_prevalence[country][dis]-mu)/std\n",
    "        country_proximal_zscore_list.append(z)\n",
    "        from scipy.stats import norm\n",
    "        ptemp = norm.sf(abs(z))\n",
    "    if len(country_proximal_zscore_list)>0:\n",
    "        country_proximal_zscore_dict[country]=country_proximal_zscore_list\n",
    "\n",
    "air_country_distant_zscore_dict={}\n",
    "for country,dis_dict in new_high_country_distant_dis_dict.items():\n",
    "    country_distant_zscore_list=[]\n",
    "    for dis in list(dis_dict.keys()):\n",
    "        mu = np.mean(distant_dis_prevalence[dis])\n",
    "        std = np.std(distant_dis_prevalence[dis])\n",
    "        z = (country_dis_prevalence[country][dis]-mu)/std\n",
    "        country_distant_zscore_list.append(z)\n",
    "        from scipy.stats import norm\n",
    "        ptemp = norm.sf(abs(z))\n",
    "    air_country_distant_zscore_dict[country]=country_distant_zscore_list\n",
    "\n",
    "air_pollution_zscore_df = pd.DataFrame()\n",
    "selected_countries=set(country_proximal_zscore_dict.keys())\n",
    "tot_country_list=[]\n",
    "tot_zscore_list=[]\n",
    "location_network_list=[]\n",
    "for country in selected_countries:\n",
    "    farawaylist_cleaned = [x for x in air_country_distant_zscore_dict[country] if str(x) != 'nan']\n",
    "    proximallist_cleaned = [x for x in country_proximal_zscore_dict[country] if str(x) != 'nan']\n",
    "    tot_zscorelist_partial=farawaylist_cleaned+proximallist_cleaned\n",
    "    for i in proximallist_cleaned:\n",
    "        location_network_list.append(\"proximal\")\n",
    "        tot_zscore_list.append(i)\n",
    "        tot_country_list.append(country)\n",
    "    for i in farawaylist_cleaned:\n",
    "        location_network_list.append(\"distant\")\n",
    "        tot_zscore_list.append(i)\n",
    "        tot_country_list.append(country)\n",
    "\n",
    "air_pollution_zscore_df['location']=tot_country_list\n",
    "air_pollution_zscore_df['z-score']=tot_zscore_list\n",
    "air_pollution_zscore_df['network_location']=location_network_list\n",
    "\n",
    "my_order = sorted(list(set(air_pollution_zscore_df['location'])))\n",
    "\n",
    "for country in country_proximal_zscore_dict.keys():\n",
    "    try:\n",
    "        proximallist_cleaned = [x for x in country_proximal_zscore_dict[country] if str(x) != 'nan']\n",
    "        distantlist_cleaned = [x for x in air_country_distant_zscore_dict[country] if str(x) != 'nan']\n",
    "        prox_mean = np.mean(proximallist_cleaned)\n",
    "        distant_mean = np.mean(distantlist_cleaned)\n",
    "        pval_noprox = scipy.stats.mannwhitneyu(proximallist_cleaned, distantlist_cleaned, use_continuity=True, alternative='greater')[1]\n",
    "        print(country,prox_mean,distant_mean,round(pval_noprox,4),round(pval_distant,4))\n",
    "    except:\n",
    "        print(country,prox_mean,distant_mean,round(pval_noprox,4))\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "0c1128a9-0584-4fc4-ac75-b47fd3911e70",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "fig, ax = plt.subplots(figsize=(15, 20))\n",
    "sns.set(style=\"white\")\n",
    "g = sns.boxplot(x=\"z-score\", y=\"location\", orient=\"h\",\n",
    "                hue=\"network_location\", palette=[\"#2B7A98\", \"#54CAF2\"], order=my_order,\n",
    "                data=air_pollution_zscore_df)\n",
    "fontsize = 36\n",
    "plt.legend(loc=\"upper right\", frameon=True, fontsize=fontsize)\n",
    "plt.yticks(fontsize=36)\n",
    "\n",
    "for ax in plt.gcf().axes:\n",
    "    l = ax.get_xlabel()\n",
    "    ax.set_xlabel(l, fontsize=36)\n",
    "    ax.set_ylabel(l, fontsize=36)\n",
    "\n",
    "plt.ylabel('', fontsize=25)\n",
    "plt.xlabel('', fontsize=25)\n",
    "ax.spines['top'].set_visible(False)\n",
    "ax.spines['right'].set_visible(False)\n",
    "ax.spines['bottom'].set_visible(True)\n",
    "ax.spines['left'].set_visible(True)\n",
    "ax.set_xticklabels(ax.get_xticks(), size=36)\n",
    "\n",
    "pv_list = []  \n",
    "locations = []  \n",
    "stars = []  \n",
    "\n",
    "for country in my_order:\n",
    "    proximallist_cleaned = [x for x in country_proximal_zscore_dict[country] if str(x) != 'nan']\n",
    "    distantlist_cleaned = [x for x in air_country_distant_zscore_dict[country] if str(x) != 'nan']\n",
    "    prox_mean = np.mean(proximallist_cleaned)\n",
    "    distant_mean = np.mean(distantlist_cleaned)\n",
    "    pval_noprox = scipy.stats.mannwhitneyu(proximallist_cleaned, distantlist_cleaned, use_continuity=True, alternative='greater')[1]\n",
    "    pv_list.append(pval_noprox)\n",
    "    locations.append(country)\n",
    "    if pval_noprox<0.001:\n",
    "        stars.append(\"***\")\n",
    "    elif pval_noprox>0.001 and pval_noprox<0.01:\n",
    "        stars.append(\"**\")\n",
    "    elif pval_noprox>0.01 and pval_noprox<0.05:\n",
    "        stars.append(\"*\")\n",
    "    else:\n",
    "        stars.append(\"n.s.\")\n",
    "\n",
    "for pv, loc, star in zip(pv_list, locations, stars):\n",
    "    yloc = my_order.index(loc) + 0.4 \n",
    "    xloc = max(air_pollution_zscore_df[air_pollution_zscore_df['location'] == loc]['z-score']) + 0.5  \n",
    "    ax.text(xloc, yloc, star, fontsize=fontsize, color=\"black\", ha='center')\n",
    "\n",
    "plt.savefig('output/Figures/SF12C_v01_post_rev.pdf',format='pdf',dpi=300, bbox_inches = \"tight\")\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "42d93d05-0b7e-47bb-b21e-494867e7414c",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "a21cb602-c1ab-43aa-a976-d9b5d34640f9",
   "metadata": {},
   "outputs": [],
   "source": [
    "#mortality\n",
    "\n",
    "country_proximal_zscore_dict={}\n",
    "for country,dis_dict in high_country_selected_chem_id_gbd_mean_dict.items():\n",
    "    country_proximal_zscore_list=[]\n",
    "    for dis in list(dis_dict.keys()):\n",
    "        mu = np.mean(proximal_dis_mortality[dis])\n",
    "        std = np.std(proximal_dis_mortality[dis])\n",
    "        z = (country_dis_mortality[country][dis]-mu)/std\n",
    "        country_proximal_zscore_list.append(z)\n",
    "        from scipy.stats import norm\n",
    "        ptemp = norm.sf(abs(z))\n",
    "    if len(country_proximal_zscore_list)>0:\n",
    "        country_proximal_zscore_dict[country]=country_proximal_zscore_list\n",
    "\n",
    "air_country_distant_zscore_dict={}\n",
    "for country,dis_dict in new_high_country_distant_dis_dict.items():\n",
    "    country_distant_zscore_list=[]\n",
    "    for dis in list(dis_dict.keys()):\n",
    "        mu = np.mean(distant_dis_mortality[dis])\n",
    "        std = np.std(distant_dis_mortality[dis])\n",
    "        z = (country_dis_mortality[country][dis]-mu)/std\n",
    "        country_distant_zscore_list.append(z)\n",
    "        from scipy.stats import norm\n",
    "        ptemp = norm.sf(abs(z))\n",
    "    air_country_distant_zscore_dict[country]=country_distant_zscore_list\n",
    "\n",
    "air_pollution_zscore_df = pd.DataFrame()\n",
    "selected_countries=set(country_proximal_zscore_dict.keys())\n",
    "tot_country_list=[]\n",
    "tot_zscore_list=[]\n",
    "location_network_list=[]\n",
    "for country in selected_countries:\n",
    "    farawaylist_cleaned = [x for x in air_country_distant_zscore_dict[country] if str(x) != 'nan']\n",
    "    proximallist_cleaned = [x for x in country_proximal_zscore_dict[country] if str(x) != 'nan']\n",
    "    tot_zscorelist_partial=farawaylist_cleaned+proximallist_cleaned\n",
    "    for i in proximallist_cleaned:\n",
    "        location_network_list.append(\"proximal\")\n",
    "        tot_zscore_list.append(i)\n",
    "        tot_country_list.append(country)\n",
    "    for i in farawaylist_cleaned:\n",
    "        location_network_list.append(\"distant\")\n",
    "        tot_zscore_list.append(i)\n",
    "        tot_country_list.append(country)\n",
    "\n",
    "air_pollution_zscore_df['location']=tot_country_list\n",
    "air_pollution_zscore_df['z-score']=tot_zscore_list\n",
    "air_pollution_zscore_df['network_location']=location_network_list\n",
    "\n",
    "my_order = sorted(list(set(air_pollution_zscore_df['location'])))\n",
    "\n",
    "for country in country_proximal_zscore_dict.keys():\n",
    "    try:\n",
    "        proximallist_cleaned = [x for x in country_proximal_zscore_dict[country] if str(x) != 'nan']\n",
    "        distantlist_cleaned = [x for x in air_country_distant_zscore_dict[country] if str(x) != 'nan']\n",
    "        prox_mean = np.mean(proximallist_cleaned)\n",
    "        distant_mean = np.mean(distantlist_cleaned)\n",
    "        pval_noprox = scipy.stats.mannwhitneyu(proximallist_cleaned, distantlist_cleaned, use_continuity=True, alternative='greater')[1]\n",
    "        print(country,prox_mean,distant_mean,round(pval_noprox,4),round(pval_distant,4))\n",
    "    except:\n",
    "        print(country,prox_mean,distant_mean,round(pval_noprox,4))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "e60ea095-54dc-4367-a22a-c8f6159110dc",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "fig, ax = plt.subplots(figsize=(15, 20))\n",
    "sns.set(style=\"white\")\n",
    "g = sns.boxplot(x=\"z-score\", y=\"location\", orient=\"h\",\n",
    "                hue=\"network_location\", palette=[\"#2B7A98\", \"#54CAF2\"], order=my_order,\n",
    "                data=air_pollution_zscore_df)\n",
    "fontsize = 36\n",
    "plt.legend(loc=\"upper right\", frameon=True, fontsize=fontsize)\n",
    "plt.yticks(fontsize=36)\n",
    "\n",
    "for ax in plt.gcf().axes:\n",
    "    l = ax.get_xlabel()\n",
    "    ax.set_xlabel(l, fontsize=36)\n",
    "    ax.set_ylabel(l, fontsize=36)\n",
    "\n",
    "plt.ylabel('', fontsize=25)\n",
    "plt.xlabel('', fontsize=25)\n",
    "ax.spines['top'].set_visible(False)\n",
    "ax.spines['right'].set_visible(False)\n",
    "ax.spines['bottom'].set_visible(True)\n",
    "ax.spines['left'].set_visible(True)\n",
    "ax.set_xticklabels(ax.get_xticks(), size=36)\n",
    "\n",
    "pv_list = []  \n",
    "locations = []  \n",
    "stars = []  \n",
    "\n",
    "for country in my_order:\n",
    "    proximallist_cleaned = [x for x in country_proximal_zscore_dict[country] if str(x) != 'nan']\n",
    "    distantlist_cleaned = [x for x in air_country_distant_zscore_dict[country] if str(x) != 'nan']\n",
    "    prox_mean = np.mean(proximallist_cleaned)\n",
    "    distant_mean = np.mean(distantlist_cleaned)\n",
    "    pval_noprox = scipy.stats.mannwhitneyu(proximallist_cleaned, distantlist_cleaned, use_continuity=True, alternative='greater')[1]\n",
    "    pv_list.append(pval_noprox)\n",
    "    locations.append(country)\n",
    "    if pval_noprox<0.001:\n",
    "        stars.append(\"***\")\n",
    "    elif pval_noprox>0.001 and pval_noprox<0.01:\n",
    "        stars.append(\"**\")\n",
    "    elif pval_noprox>0.01 and pval_noprox<0.05:\n",
    "        stars.append(\"*\")\n",
    "    else:\n",
    "        stars.append(\"n.s.\")\n",
    "\n",
    "for pv, loc, star in zip(pv_list, locations, stars):\n",
    "    yloc = my_order.index(loc) + 0.4 \n",
    "    xloc = max(air_pollution_zscore_df[air_pollution_zscore_df['location'] == loc]['z-score']) + 0.5  \n",
    "    ax.text(xloc, yloc, star, fontsize=fontsize, color=\"black\", ha='center')\n",
    "\n",
    "plt.savefig('output/Figures/SF13C_v01_post_rev.pdf',format='pdf',dpi=300, bbox_inches = \"tight\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "6e618c21-68f8-4cdf-a662-5ed5069edddc",
   "metadata": {},
   "outputs": [],
   "source": [
    "#daly\n",
    "\n",
    "country_proximal_zscore_dict={}\n",
    "for country,dis_dict in high_country_selected_chem_id_gbd_mean_dict.items():\n",
    "    country_proximal_zscore_list=[]\n",
    "    for dis in list(dis_dict.keys()):\n",
    "        mu = np.mean(proximal_dis_daly[dis])\n",
    "        std = np.std(proximal_dis_daly[dis])\n",
    "        z = (country_dis_DALYs[country][dis]-mu)/std\n",
    "        country_proximal_zscore_list.append(z)\n",
    "        from scipy.stats import norm\n",
    "        ptemp = norm.sf(abs(z))\n",
    "    if len(country_proximal_zscore_list)>0:\n",
    "        country_proximal_zscore_dict[country]=country_proximal_zscore_list\n",
    "\n",
    "air_country_distant_zscore_dict={}\n",
    "for country,dis_dict in new_high_country_distant_dis_dict.items():\n",
    "    country_distant_zscore_list=[]\n",
    "    for dis in list(dis_dict.keys()):\n",
    "        mu = np.mean(distant_dis_daly[dis])\n",
    "        std = np.std(distant_dis_daly[dis])\n",
    "        z = (country_dis_DALYs[country][dis]-mu)/std\n",
    "        country_distant_zscore_list.append(z)\n",
    "        from scipy.stats import norm\n",
    "        ptemp = norm.sf(abs(z))\n",
    "    air_country_distant_zscore_dict[country]=country_distant_zscore_list\n",
    "\n",
    "air_pollution_zscore_df = pd.DataFrame()\n",
    "selected_countries=set(country_proximal_zscore_dict.keys())\n",
    "tot_country_list=[]\n",
    "tot_zscore_list=[]\n",
    "location_network_list=[]\n",
    "for country in selected_countries:\n",
    "    farawaylist_cleaned = [x for x in air_country_distant_zscore_dict[country] if str(x) != 'nan']\n",
    "    proximallist_cleaned = [x for x in country_proximal_zscore_dict[country] if str(x) != 'nan']\n",
    "    tot_zscorelist_partial=farawaylist_cleaned+proximallist_cleaned\n",
    "    for i in proximallist_cleaned:\n",
    "        location_network_list.append(\"proximal\")\n",
    "        tot_zscore_list.append(i)\n",
    "        tot_country_list.append(country)\n",
    "    for i in farawaylist_cleaned:\n",
    "        location_network_list.append(\"distant\")\n",
    "        tot_zscore_list.append(i)\n",
    "        tot_country_list.append(country)\n",
    "\n",
    "air_pollution_zscore_df['location']=tot_country_list\n",
    "air_pollution_zscore_df['z-score']=tot_zscore_list\n",
    "air_pollution_zscore_df['network_location']=location_network_list\n",
    "\n",
    "my_order = sorted(list(set(air_pollution_zscore_df['location'])))\n",
    "\n",
    "for country in country_proximal_zscore_dict.keys():\n",
    "    try:\n",
    "        proximallist_cleaned = [x for x in country_proximal_zscore_dict[country] if str(x) != 'nan']\n",
    "        distantlist_cleaned = [x for x in air_country_distant_zscore_dict[country] if str(x) != 'nan']\n",
    "        prox_mean = np.mean(proximallist_cleaned)\n",
    "        distant_mean = np.mean(distantlist_cleaned)\n",
    "        pval_noprox = scipy.stats.mannwhitneyu(proximallist_cleaned, distantlist_cleaned, use_continuity=True, alternative='greater')[1]\n",
    "        print(country,prox_mean,distant_mean,round(pval_noprox,4),round(pval_distant,4))\n",
    "    except:\n",
    "        print(country,prox_mean,distant_mean,round(pval_noprox,4))\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c12cf15a-6b03-48bc-bed4-060fca46d01d",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "fig, ax = plt.subplots(figsize=(15, 20))\n",
    "sns.set(style=\"white\")\n",
    "g = sns.boxplot(x=\"z-score\", y=\"location\", orient=\"h\",\n",
    "                hue=\"network_location\", palette=[\"#2B7A98\", \"#54CAF2\"], order=my_order,\n",
    "                data=air_pollution_zscore_df)\n",
    "fontsize = 36\n",
    "plt.legend(loc=\"upper right\", frameon=True, fontsize=fontsize)\n",
    "plt.yticks(fontsize=36)\n",
    "\n",
    "for ax in plt.gcf().axes:\n",
    "    l = ax.get_xlabel()\n",
    "    ax.set_xlabel(l, fontsize=36)\n",
    "    ax.set_ylabel(l, fontsize=36)\n",
    "\n",
    "plt.ylabel('', fontsize=25)\n",
    "plt.xlabel('', fontsize=25)\n",
    "ax.spines['top'].set_visible(False)\n",
    "ax.spines['right'].set_visible(False)\n",
    "ax.spines['bottom'].set_visible(True)\n",
    "ax.spines['left'].set_visible(True)\n",
    "ax.set_xticklabels(ax.get_xticks(), size=36)\n",
    "\n",
    "pv_list = []  \n",
    "locations = []  \n",
    "stars = []  \n",
    "\n",
    "for country in my_order:\n",
    "    proximallist_cleaned = [x for x in country_proximal_zscore_dict[country] if str(x) != 'nan']\n",
    "    distantlist_cleaned = [x for x in air_country_distant_zscore_dict[country] if str(x) != 'nan']\n",
    "    prox_mean = np.mean(proximallist_cleaned)\n",
    "    distant_mean = np.mean(distantlist_cleaned)\n",
    "    pval_noprox = scipy.stats.mannwhitneyu(proximallist_cleaned, distantlist_cleaned, use_continuity=True, alternative='greater')[1]\n",
    "    pv_list.append(pval_noprox)\n",
    "    locations.append(country)\n",
    "    if pval_noprox<0.001:\n",
    "        stars.append(\"***\")\n",
    "    elif pval_noprox>0.001 and pval_noprox<0.01:\n",
    "        stars.append(\"**\")\n",
    "    elif pval_noprox>0.01 and pval_noprox<0.05:\n",
    "        stars.append(\"*\")\n",
    "    else:\n",
    "        stars.append(\"n.s.\")\n",
    "\n",
    "for pv, loc, star in zip(pv_list, locations, stars):\n",
    "    yloc = my_order.index(loc) + 0.4 \n",
    "    xloc = max(air_pollution_zscore_df[air_pollution_zscore_df['location'] == loc]['z-score']) + 0.5  \n",
    "    ax.text(xloc, yloc, star, fontsize=fontsize, color=\"black\", ha='center')\n",
    "\n",
    "plt.savefig('output/Figures/SF14C_v01_post_rev.pdf',format='pdf',dpi=300, bbox_inches = \"tight\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d8bda47d-606b-423a-a428-5b4a502a892c",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "8ef8c01a-643c-437e-a244-d3d80e5a4e16",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "003a867c-d19c-4548-9134-80c7657689e9",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d5ba01e5-b3de-4d7f-913e-90317975f577",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "16e49a4b-51ad-40fa-8008-d5fdb23d80d9",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "7f372a0c-65a4-4b3d-b887-4bdabd71eb0b",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "id": "fbea01ec-b7d4-48eb-aaec-c182b0d209c1",
   "metadata": {},
   "source": [
    "### Water"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d6e6e874-1a9f-424d-979d-7af13a7e1fe0",
   "metadata": {},
   "outputs": [],
   "source": [
    "water_chem_name_conversion_dict = air_chem_name_conversion_dict.copy()\n",
    "\n",
    "#Let's add the chemicals that are only in the water and not in the air\n",
    "water_chem_name_conversion_dict['Triphenyltin and compounds'] = 'C030665'\n",
    "water_chem_name_conversion_dict['Isoproturon'] = 'C028904'\n",
    "water_chem_name_conversion_dict['Mirex'] = 'D008917'\n",
    "water_chem_name_conversion_dict['Hexachlorobutadiene'] = 'C001335'\n",
    "water_chem_name_conversion_dict['Heptachlor'] = 'D006533'\n",
    "\n",
    "water_chem_name_conversion_dict['Diuron'] = 'D004237'\n",
    "water_chem_name_conversion_dict['Chlorpyrifos'] = 'D004390'\n",
    "water_chem_name_conversion_dict['Fluoranthene'] = 'C007738'\n",
    "water_chem_name_conversion_dict['Toxaphene'] = 'D014112'\n",
    "water_chem_name_conversion_dict['Halogenated organic compounds (as AOX)'] = 'D006846'\n",
    "\n",
    "water_chem_name_conversion_dict['Simazine'] = 'D012839'\n",
    "water_chem_name_conversion_dict['Atrazine'] = 'D001280'\n",
    "water_chem_name_conversion_dict['Total nitrogen'] = 'D009584'\n",
    "water_chem_name_conversion_dict['Dieldrin'] = 'D004026'\n",
    "water_chem_name_conversion_dict['Chlordane'] = 'D002706'\n",
    "\n",
    "\n",
    "water_chem_name_conversion_dict['Alachlor'] = 'C000188'\n",
    "water_chem_name_conversion_dict['Chlorides'] = 'D002712'\n",
    "water_chem_name_conversion_dict['DDT (DDT)'] = 'D003634'\n",
    "water_chem_name_conversion_dict['Tributyltin and compounds'] = 'C011559'\n",
    "water_chem_name_conversion_dict['Chlorfenvinphos'] = 'D002709'\n",
    "\n",
    "\n",
    "water_chem_name_conversion_dict['Endrin'] = 'D004732'\n",
    "water_chem_name_conversion_dict['Organotin compounds (as total Sn)'] = 'D009947'\n",
    "water_chem_name_conversion_dict['1,2,3,4,5,6-hexachlorocyclohexane (HCH)'] = 'D001556'\n",
    "water_chem_name_conversion_dict['Trifluralin'] = 'D014274'\n",
    "water_chem_name_conversion_dict['Cyanides (as total CN)'] = 'D003486'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "dd945729-a69f-453a-b2c3-21c534a2bfd6",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Let's import the WHO data at a national level\n",
    "#water\n",
    "who_water_nation=pd.read_excel(\"input/WHO/F2_1_Total Releases at National Level into Water.xlsx\",sheet_name=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "4bfbc934-d2ed-4d0d-b9e9-d8d4c89f7d53",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "country_set=set(who_water_nation['countryName'].tolist())\n",
    "years=list(range(2007,2020))\n",
    "country_chem_conc_water={}\n",
    "for c in country_set:\n",
    "    chem_concentration_dict={}\n",
    "    for i,v in who_water_nation[who_water_nation['countryName']==c].iterrows():\n",
    "        concentration=[]\n",
    "        for y in years:\n",
    "            concentration.append(v[y])\n",
    "        concentration_clenaed=[x for x in concentration if str(x) != 'nan']\n",
    "        chem_concentration_dict[v['pollutant']]=concentration_clenaed\n",
    "    country_chem_conc_water[c]=chem_concentration_dict"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "7176ba83-c951-470e-bd8c-a9d0c68c0e0c",
   "metadata": {},
   "outputs": [],
   "source": [
    "country_chem_conc_mean_water_dict={}\n",
    "for c,chem_con in country_chem_conc_water.items():\n",
    "    chem_mean_dict={}\n",
    "    for chem,con_list in chem_con.items():\n",
    "        chem_mean_dict[chem]=mean_list(con_list)\n",
    "    country_chem_conc_mean_water_dict[c]=chem_mean_dict"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "4e2bfb2d-2505-4229-aaa4-52f6e2e6b70d",
   "metadata": {},
   "outputs": [],
   "source": [
    "chem_set=set(who_water_nation['pollutant'].tolist())\n",
    "\n",
    "chem_across_countries_water_dict={}\n",
    "for chem in chem_set:\n",
    "    chem_median_across_countries_list=[]\n",
    "    for c in country_set:\n",
    "        try:\n",
    "            chem_median_across_countries_list.append(country_chem_conc_mean_water_dict[c][chem])\n",
    "        except:\n",
    "            pass\n",
    "    chem_across_countries_water_dict[chem]=chem_median_across_countries_list\n",
    "    \n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "f149dcb5-6ab7-4740-8591-7af9d8ad72ef",
   "metadata": {},
   "outputs": [],
   "source": [
    "#If a chemical compound is express only in less than 4 countries then is considered as specific for that country. \n",
    "#Otherwise the concentration in that country should be > 75* percentile of the average concentration distribution across all countries\n",
    "import statistics\n",
    "\n",
    "water_country_selected_chem_dict={}\n",
    "\n",
    "for country, chem_avgcon_dict in country_chem_conc_mean_water_dict.items():\n",
    "    chem_list=[]\n",
    "    for chem, avg_con_list in chem_across_countries_water_dict.items():\n",
    "        if chem in chem_avgcon_dict.keys():\n",
    "            if len(avg_con_list)>4:\n",
    "                cutoff=np.percentile(avg_con_list,75)\n",
    "                if chem_avgcon_dict[chem]>cutoff:\n",
    "                    chem_list.append(chem)\n",
    "            else:\n",
    "                chem_list.append(chem)\n",
    "    water_country_selected_chem_dict[country]=chem_list\n",
    "\n",
    "\n",
    "water_low_country_selected_chem_dict={}\n",
    "\n",
    "for country, chem_avgcon_dict in country_chem_conc_mean_water_dict.items():\n",
    "    chem_list=[]\n",
    "    for chem, avg_con_list in chem_across_countries_water_dict.items():\n",
    "        if chem in chem_avgcon_dict.keys():\n",
    "            if len(avg_con_list)>4:\n",
    "                cutoff=np.percentile(avg_con_list,50)\n",
    "                if chem_avgcon_dict[chem]>cutoff:\n",
    "                    chem_list.append(chem)\n",
    "            else:\n",
    "                chem_list.append(chem)\n",
    "    water_low_country_selected_chem_dict[country]=chem_list\n",
    "\n",
    "water_high_country_selected_chem_dict={}\n",
    "\n",
    "for country, chem_avgcon_dict in country_chem_conc_mean_water_dict.items():\n",
    "    chem_list=[]\n",
    "    for chem, avg_con_list in chem_across_countries_water_dict.items():\n",
    "        if chem in chem_avgcon_dict.keys():\n",
    "            if len(avg_con_list)>4:\n",
    "                cutoff=np.percentile(avg_con_list,95)\n",
    "                if chem_avgcon_dict[chem]>cutoff:\n",
    "                    chem_list.append(chem)\n",
    "            else:\n",
    "                chem_list.append(chem)\n",
    "    water_high_country_selected_chem_dict[country]=chem_list"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "6f3e38cb-787c-48bf-81cc-66b4e229cde7",
   "metadata": {},
   "outputs": [],
   "source": [
    "water_country_selected_chem_id_dict={}\n",
    "for country,chem_list in water_country_selected_chem_dict.items():\n",
    "    id_chem_list=[]\n",
    "    for chem in chem_list:\n",
    "        try:\n",
    "            if chem!='PCDD + PCDF (dioxins + furans) (as Teq)':   \n",
    "                if water_chem_name_conversion_dict[chem] in chem_id_list:\n",
    "                    id_chem_list.append(water_chem_name_conversion_dict[chem])\n",
    "            else:\n",
    "                if 'D005663' in chem_id_list:\n",
    "                    id_chem_list.append('D005663')\n",
    "                if 'D000072316' in chem_id_list:\n",
    "                    id_chem_list.append('D000072316')\n",
    "                if 'D004147' in chem_id_list:\n",
    "                    id_chem_list.append('D004147')\n",
    "        except:\n",
    "            pass\n",
    "    water_country_selected_chem_id_dict[country]=set(id_chem_list)\n",
    "\n",
    "water_low_country_selected_chem_id_dict={}\n",
    "for country,chem_list in water_low_country_selected_chem_dict.items():\n",
    "    id_chem_list=[]\n",
    "    for chem in chem_list:\n",
    "        try:\n",
    "            if chem!='PCDD + PCDF (dioxins + furans) (as Teq)':   \n",
    "                if water_chem_name_conversion_dict[chem] in chem_id_list:\n",
    "                    id_chem_list.append(water_chem_name_conversion_dict[chem])\n",
    "            else:\n",
    "                if 'D005663' in chem_id_list:\n",
    "                    id_chem_list.append('D005663')\n",
    "                if 'D000072316' in chem_id_list:\n",
    "                    id_chem_list.append('D000072316')\n",
    "                if 'D004147' in chem_id_list:\n",
    "                    id_chem_list.append('D004147')\n",
    "        except:\n",
    "            pass\n",
    "    water_low_country_selected_chem_id_dict[country]=set(id_chem_list)\n",
    "\n",
    "water_high_country_selected_chem_id_dict={}\n",
    "for country,chem_list in water_high_country_selected_chem_dict.items():\n",
    "    id_chem_list=[]\n",
    "    for chem in chem_list:\n",
    "        try:\n",
    "            if chem!='PCDD + PCDF (dioxins + furans) (as Teq)':   \n",
    "                if water_chem_name_conversion_dict[chem] in chem_id_list:\n",
    "                    id_chem_list.append(water_chem_name_conversion_dict[chem])\n",
    "            else:\n",
    "                if 'D005663' in chem_id_list:\n",
    "                    id_chem_list.append('D005663')\n",
    "                if 'D000072316' in chem_id_list:\n",
    "                    id_chem_list.append('D000072316')\n",
    "                if 'D004147' in chem_id_list:\n",
    "                    id_chem_list.append('D004147')\n",
    "        except:\n",
    "            pass\n",
    "    water_high_country_selected_chem_id_dict[country]=set(id_chem_list)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ca553575-6859-43f0-ab56-11c7afccf758",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "water_low_country_selected_chem_num_dict = {}\n",
    "for country, chemlist in water_low_country_selected_chem_id_dict.items():\n",
    "    water_low_country_selected_chem_num_dict[country] = len(chemlist)\n",
    "\n",
    "water_country_selected_chem_num_dict = {}\n",
    "for country, chemlist in water_country_selected_chem_id_dict.items():\n",
    "    water_country_selected_chem_num_dict[country] = len(chemlist)\n",
    "\n",
    "water_high_country_selected_chem_num_dict = {}\n",
    "for country, chemlist in water_high_country_selected_chem_id_dict.items():\n",
    "    water_high_country_selected_chem_num_dict[country] = len(chemlist)\n",
    "    \n",
    "# Get the list of countries (keys)\n",
    "countries = list(water_country_selected_chem_num_dict.keys())\n",
    "\n",
    "# Extract values from each dictionary\n",
    "values_a = [water_low_country_selected_chem_num_dict[country] for country in countries]\n",
    "values_b = [water_country_selected_chem_num_dict[country] for country in countries]\n",
    "values_c = [water_high_country_selected_chem_num_dict[country] for country in countries]\n",
    "\n",
    "# Define the position of bars on the x-axis\n",
    "x = np.arange(len(countries))\n",
    "\n",
    "# Width of a bar\n",
    "width = 0.2\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(15, 20))\n",
    "\n",
    "# Plotting the bars\n",
    "rects1 = ax.barh(x - width, values_a, width, color='blue', label='50th')\n",
    "rects2 = ax.barh(x, values_b, width, color='black', label='75th')\n",
    "rects3 = ax.barh(x + width, values_c, width, color='red', label='95th')\n",
    "\n",
    "# Adding labels, title, and customizing axes\n",
    "ax.set_xlabel('Number of exposures')\n",
    "ax.set_ylabel('Countries')\n",
    "ax.set_title('Number of exposures per country')\n",
    "ax.set_yticks(x)\n",
    "ax.set_yticklabels(countries,fontsize=24)\n",
    "ax.legend()\n",
    "\n",
    "# Show the plot\n",
    "plt.tight_layout()\n",
    "plt.savefig('output/Figures/SF8E_v01_post_rev.pdf',format='pdf',dpi=300, bbox_inches = \"tight\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "40d6c685-ed58-4446-be17-2ba58a42a3c6",
   "metadata": {},
   "outputs": [],
   "source": [
    "print(np.mean(list(water_low_country_selected_chem_num_dict.values())))\n",
    "print(np.mean(list(water_country_selected_chem_num_dict.values())))\n",
    "print(np.mean(list(water_high_country_selected_chem_num_dict.values())))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "3c6696a0-f56e-4746-81f3-9b8aa390bc13",
   "metadata": {},
   "outputs": [],
   "source": [
    "jaccard_water_country_selected_chem_dict = {}\n",
    "\n",
    "for country1,chemlist1 in water_country_selected_chem_id_dict.items():\n",
    "    for country2,chemlist2 in water_country_selected_chem_id_dict.items():\n",
    "        if len(chemlist1)>0 and len(chemlist1)>0:\n",
    "            jaccard_water_country_selected_chem_dict[country1,country2] = overlap_jaccard(chemlist1,chemlist2)\n",
    "        else:\n",
    "            jaccard_water_country_selected_chem_dict[country1,country2] = 0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "77feb54c-4434-41f6-9461-b6a45ff49569",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "ser = pd.Series(list(jaccard_water_country_selected_chem_dict.values()),\n",
    "                index=pd.MultiIndex.from_tuples(jaccard_water_country_selected_chem_dict.keys()))\n",
    "jaccard_water_country_selected_chem_dict_df = ser.unstack().fillna(0)\n",
    "\n",
    "# Create a mask for the higher triangle\n",
    "mask = np.triu(np.ones_like(jaccard_water_country_selected_chem_dict_df, dtype=bool))\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(24, 18))\n",
    "\n",
    "# Plot heatmap with the mask\n",
    "hm = sns.heatmap(jaccard_water_country_selected_chem_dict_df, cbar=True, mask=mask, vmin=0, vmax=1,\n",
    "                 fmt='.2f', annot_kws={'size': 10}, annot=True, \n",
    "                 square=True, cmap='YlGnBu')\n",
    "\n",
    "ticks = np.arange(jaccard_water_country_selected_chem_dict_df.shape[0]) + 0.5\n",
    "ax.set_xticks(ticks)\n",
    "ax.set_xticklabels(jaccard_water_country_selected_chem_dict_df.columns, rotation=90, fontsize=24)\n",
    "ax.set_yticks(ticks)\n",
    "ax.set_yticklabels(jaccard_water_country_selected_chem_dict_df.index, rotation=360, fontsize=24)\n",
    "\n",
    "ax.set_title('Jaccard index')\n",
    "plt.tight_layout()\n",
    "plt.savefig('output/Figures/SD9D_v01_post_rev.pdf',format='pdf',dpi=300, bbox_inches = \"tight\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "5c2a8c86-9ee8-47b7-9574-d67225d7ad72",
   "metadata": {},
   "outputs": [],
   "source": [
    "water_country_selected_chem_dict_ji_list = []\n",
    "for country1,chemlist1 in water_country_selected_chem_id_dict.items():\n",
    "    for country2,chemlist2 in water_country_selected_chem_id_dict.items():\n",
    "        if len(chemlist1)>0 and len(chemlist1)>0:\n",
    "            if country1!=country2:\n",
    "                water_country_selected_chem_dict_ji_list.append(overlap_jaccard(chemlist1,chemlist2))\n",
    "print(np.mean(water_country_selected_chem_dict_ji_list))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "0836f2fa-9080-45dd-ac5d-ac008d3661c6",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "9ecfb33f-1787-45c8-800d-235e89fd9c74",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "7ea18a08-8d6a-477c-a558-b99445caf3ad",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "15e28154-acf6-414b-a8e1-d1656a54a563",
   "metadata": {},
   "outputs": [],
   "source": [
    "jaccard_water_low_country_selected_chem_dict = {}\n",
    "\n",
    "for country1,chemlist1 in water_low_country_selected_chem_id_dict.items():\n",
    "    for country2,chemlist2 in water_low_country_selected_chem_id_dict.items():\n",
    "        if len(chemlist1)>0 and len(chemlist1)>0:\n",
    "            jaccard_water_low_country_selected_chem_dict[country1,country2] = overlap_jaccard(chemlist1,chemlist2)\n",
    "        else:\n",
    "            jaccard_water_low_country_selected_chem_dict[country1,country2] = 0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "2e54a688-9cdf-4685-ac1f-9156e07fb0a7",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "ser = pd.Series(list(jaccard_water_low_country_selected_chem_dict.values()),\n",
    "                index=pd.MultiIndex.from_tuples(jaccard_water_low_country_selected_chem_dict.keys()))\n",
    "jaccard_water_low_country_selected_chem_dict_df = ser.unstack().fillna(0)\n",
    "\n",
    "# Create a mask for the higher triangle\n",
    "mask = np.triu(np.ones_like(jaccard_water_low_country_selected_chem_dict_df, dtype=bool))\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(24, 18))\n",
    "\n",
    "# Plot heatmap with the mask\n",
    "hm = sns.heatmap(jaccard_water_low_country_selected_chem_dict_df, cbar=True, mask=mask, vmin=0, vmax=1,\n",
    "                 fmt='.2f', annot_kws={'size': 10}, annot=True, \n",
    "                 square=True, cmap='YlGnBu')\n",
    "\n",
    "ticks = np.arange(jaccard_water_low_country_selected_chem_dict_df.shape[0]) + 0.5\n",
    "ax.set_xticks(ticks)\n",
    "ax.set_xticklabels(jaccard_water_low_country_selected_chem_dict_df.columns, rotation=90, fontsize=24)\n",
    "ax.set_yticks(ticks)\n",
    "ax.set_yticklabels(jaccard_water_low_country_selected_chem_dict_df.index, rotation=360, fontsize=24)\n",
    "\n",
    "ax.set_title('Jaccard index')\n",
    "plt.tight_layout()\n",
    "plt.savefig('output/Figures/SD9E_v01_post_rev.pdf',format='pdf',dpi=300, bbox_inches = \"tight\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d2b8bf93-a470-4d72-94dd-d6e47440c3e2",
   "metadata": {},
   "outputs": [],
   "source": [
    "water_low_country_selected_chem_dict_ji_list = []\n",
    "for country1,chemlist1 in water_low_country_selected_chem_id_dict.items():\n",
    "    for country2,chemlist2 in water_low_country_selected_chem_id_dict.items():\n",
    "        if len(chemlist1)>0 and len(chemlist1)>0:\n",
    "            if country1!=country2:\n",
    "                water_low_country_selected_chem_dict_ji_list.append(overlap_jaccard(chemlist1,chemlist2))\n",
    "print(np.mean(water_low_country_selected_chem_dict_ji_list))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "7fcfec83-50ef-4452-bd7c-0d360d62ea62",
   "metadata": {},
   "outputs": [],
   "source": [
    "jaccard_water_high_country_selected_chem_dict = {}\n",
    "\n",
    "for country1,chemlist1 in water_high_country_selected_chem_id_dict.items():\n",
    "    for country2,chemlist2 in water_high_country_selected_chem_id_dict.items():\n",
    "        if len(chemlist1)>0 and len(chemlist1)>0:\n",
    "            jaccard_water_high_country_selected_chem_dict[country1,country2] = overlap_jaccard(chemlist1,chemlist2)\n",
    "        else:\n",
    "            jaccard_water_high_country_selected_chem_dict[country1,country2] = 0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "35643700-cb7c-4cd1-ad4f-a56986854b82",
   "metadata": {},
   "outputs": [],
   "source": [
    "ser = pd.Series(list(jaccard_water_high_country_selected_chem_dict.values()),\n",
    "                index=pd.MultiIndex.from_tuples(jaccard_water_high_country_selected_chem_dict.keys()))\n",
    "jaccard_water_high_country_selected_chem_dict_df = ser.unstack().fillna(0)\n",
    "\n",
    "# Create a mask for the higher triangle\n",
    "mask = np.triu(np.ones_like(jaccard_water_high_country_selected_chem_dict_df, dtype=bool))\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(24, 18))\n",
    "\n",
    "# Plot heatmap with the mask\n",
    "hm = sns.heatmap(jaccard_water_high_country_selected_chem_dict_df, cbar=True, mask=mask, vmin=0, vmax=1,\n",
    "                 fmt='.2f', annot_kws={'size': 10}, annot=True, \n",
    "                 square=True, cmap='YlGnBu')\n",
    "\n",
    "ticks = np.arange(jaccard_water_high_country_selected_chem_dict_df.shape[0]) + 0.5\n",
    "ax.set_xticks(ticks)\n",
    "ax.set_xticklabels(jaccard_water_high_country_selected_chem_dict_df.columns, rotation=90, fontsize=24)\n",
    "ax.set_yticks(ticks)\n",
    "ax.set_yticklabels(jaccard_water_high_country_selected_chem_dict_df.index, rotation=360, fontsize=24)\n",
    "\n",
    "ax.set_title('Jaccard index')\n",
    "plt.tight_layout()\n",
    "plt.savefig('output/Figures/SD9F_v01_post_rev.pdf',format='pdf',dpi=300, bbox_inches = \"tight\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c3435168-0be5-4abb-bcac-96f3f956ff09",
   "metadata": {},
   "outputs": [],
   "source": [
    "water_high_country_selected_chem_dict_ji_list = []\n",
    "for country1,chemlist1 in water_high_country_selected_chem_id_dict.items():\n",
    "    for country2,chemlist2 in water_high_country_selected_chem_id_dict.items():\n",
    "        if len(chemlist1)>0 and len(chemlist1)>0:\n",
    "            if country1!=country2:\n",
    "                water_high_country_selected_chem_dict_ji_list.append(overlap_jaccard(chemlist1,chemlist2))\n",
    "print(np.mean(water_high_country_selected_chem_dict_ji_list))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "531474e9-4e7a-4bb0-899d-55e6696a92b6",
   "metadata": {},
   "outputs": [],
   "source": [
    "#DISEASES"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "f86e7347-f22d-43e3-9a4d-43e76f3e0c58",
   "metadata": {},
   "outputs": [],
   "source": [
    "water_country_selected_dis_dict={}\n",
    "for country, chem_list in water_country_selected_chem_id_dict.items():\n",
    "    dis_list=[]\n",
    "    for chem in chem_list:\n",
    "        for chem_dis_pair,val in real_significant_final_exp_proximity.items():\n",
    "            if chem in chem_dis_pair[0]:\n",
    "                dis_list.append(chem_dis_pair[1])\n",
    "    water_country_selected_dis_dict[country]=set(dis_list)\n",
    "\n",
    "water_low_country_selected_dis_dict={}\n",
    "for country, chem_list in water_low_country_selected_chem_id_dict.items():\n",
    "    dis_list=[]\n",
    "    for chem in chem_list:\n",
    "        for chem_dis_pair,val in real_significant_final_exp_proximity.items():\n",
    "            if chem in chem_dis_pair[0]:\n",
    "                dis_list.append(chem_dis_pair[1])\n",
    "    water_low_country_selected_dis_dict[country]=set(dis_list)\n",
    "\n",
    "water_high_country_selected_dis_dict={}\n",
    "for country, chem_list in water_high_country_selected_chem_id_dict.items():\n",
    "    dis_list=[]\n",
    "    for chem in chem_list:\n",
    "        for chem_dis_pair,val in real_significant_final_exp_proximity.items():\n",
    "            if chem in chem_dis_pair[0]:\n",
    "                dis_list.append(chem_dis_pair[1])\n",
    "    water_high_country_selected_dis_dict[country]=set(dis_list)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "a256f243-2d05-4a22-96bc-42b84e285e6c",
   "metadata": {},
   "outputs": [],
   "source": [
    "water_proximal_disease_list = [] \n",
    "\n",
    "for country, disset in water_country_selected_dis_dict.items():\n",
    "    for dis in disset:\n",
    "        if dis not in water_proximal_disease_list:\n",
    "            water_proximal_disease_list.append(dis)\n",
    "\n",
    "print(len(water_proximal_disease_list))\n",
    "\n",
    "low_water_proximal_disease_list = [] \n",
    "for country, disset in water_low_country_selected_dis_dict.items():\n",
    "    for dis in disset:\n",
    "        if dis not in low_water_proximal_disease_list:\n",
    "            low_water_proximal_disease_list.append(dis)\n",
    "\n",
    "print(len(low_water_proximal_disease_list))\n",
    "\n",
    "high_water_proximal_disease_list = [] \n",
    "\n",
    "for country, disset in water_high_country_selected_dis_dict.items():\n",
    "    for dis in disset:\n",
    "        if dis not in high_water_proximal_disease_list:\n",
    "            high_water_proximal_disease_list.append(dis)\n",
    "\n",
    "print(len(high_water_proximal_disease_list))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "5f6192a3-3858-4ca9-960f-698a2861779b",
   "metadata": {},
   "outputs": [],
   "source": [
    "for dis in low_water_proximal_disease_list:\n",
    "    if dis not in water_proximal_disease_list:\n",
    "        print(dis)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c836116a-44a4-45e2-882f-7efbcac8ab49",
   "metadata": {},
   "outputs": [],
   "source": [
    "for dis in high_water_proximal_disease_list:\n",
    "    if dis not in water_proximal_disease_list:\n",
    "        print(dis)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "fd19d607-d548-4174-9775-2a47af8522a7",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "\n",
    "water_low_country_selected_dis_num_dict = {}\n",
    "for country, dislist in water_low_country_selected_dis_dict.items():\n",
    "    water_low_country_selected_dis_num_dict[country] = len(dislist)\n",
    "\n",
    "water_country_selected_dis_num_dict = {}\n",
    "for country, dislist in water_country_selected_dis_dict.items():\n",
    "    water_country_selected_dis_num_dict[country] = len(dislist)\n",
    "    \n",
    "water_high_country_selected_dis_num_dict = {}\n",
    "for country, dislist in water_high_country_selected_dis_dict.items():\n",
    "    water_high_country_selected_dis_num_dict[country] = len(dislist)\n",
    "    \n",
    "# Get the list of countries (keys)\n",
    "countries = list(water_country_selected_dis_dict.keys())\n",
    "\n",
    "# Extract values from each dictionary\n",
    "values_a = [water_low_country_selected_dis_num_dict[country] for country in countries]\n",
    "values_b = [water_country_selected_dis_num_dict[country] for country in countries]\n",
    "values_c = [water_high_country_selected_dis_num_dict[country] for country in countries]\n",
    "\n",
    "# Define the position of bars on the x-axis\n",
    "x = np.arange(len(countries))\n",
    "\n",
    "# Width of a bar\n",
    "width = 0.2\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(15, 20))\n",
    "\n",
    "# Plotting the bars\n",
    "rects1 = ax.barh(x - width, values_a, width, color='blue', label='50th')\n",
    "rects2 = ax.barh(x, values_b, width, color='black', label='75th')\n",
    "rects3 = ax.barh(x + width, values_c, width, color='red', label='95th')\n",
    "\n",
    "# Adding labels, title, and customizing axes\n",
    "ax.set_xlabel('Number of proximal diseases')\n",
    "ax.set_ylabel('Countries')\n",
    "ax.set_title('Number of proximal per country')\n",
    "ax.set_yticks(x)\n",
    "ax.set_yticklabels(countries,fontsize=24)\n",
    "ax.legend()\n",
    "\n",
    "# Show the plot\n",
    "plt.tight_layout()\n",
    "plt.savefig('output/Figures/SF10B_v01_post_rev.pdf',format='pdf',dpi=300, bbox_inches = \"tight\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c41abaa5-0a84-416c-a4fa-34c6ba1933e7",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "58e459c8-b013-4437-886e-9253ffbd396d",
   "metadata": {},
   "outputs": [],
   "source": [
    "water_disgenet_gbd_dict = air_disgenet_gbd_dict.copy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "4ebf2349-d2c8-4a17-89b0-d26fb4ffacea",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Let's add some more diseases manually\n",
    "water_disgenet_gbd_dict['Amyotrophic Lateral Sclerosis']='Motor neuron disease'\n",
    "water_disgenet_gbd_dict['Familial thoracic aortic aneurysm and aortic dissection']='Aortic aneurysm'\n",
    "water_disgenet_gbd_dict['Bladder Neoplasm']='Bladder cancer'\n",
    "water_disgenet_gbd_dict['Malignant neoplasm of esophagus']='Esophageal cancer'\n",
    "water_disgenet_gbd_dict['Acute Promyelocytic Leukemia']='Leukemia'\n",
    "water_disgenet_gbd_dict['Precursor T-Cell Lymphoblastic Leukemia-Lymphoma']='Leukemia'\n",
    "water_disgenet_gbd_dict['Juvenile Myelomonocytic Leukemia']='Leukemia'\n",
    "water_disgenet_gbd_dict['MIGRAINE WITH OR WITHOUT AURA, SUSCEPTIBILITY TO, 1']='Migraine'\n",
    "water_disgenet_gbd_dict['Male infertility']='Male infertility'\n",
    "water_disgenet_gbd_dict['Precursor Cell Lymphoblastic Leukemia Lymphoma']='Leukemia'\n",
    "water_disgenet_gbd_dict['Osteoarthritis of hip']='Osteoarthritis hip'\n",
    "water_disgenet_gbd_dict['Stomach Carcinoma']='Stomach cancer'\n",
    "water_disgenet_gbd_dict['Hypertrophic Cardiomyopathy']='Other cardiomyopathy'\n",
    "water_disgenet_gbd_dict['Cerebral Hemorrhage']='Intracerebral hemorrhage'\n",
    "\n",
    "print(len(water_disgenet_gbd_dict))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "78834f01-8cc8-4db1-be0d-35d14905573a",
   "metadata": {},
   "outputs": [],
   "source": [
    "country_water_dis_mean_prox_dict = {}\n",
    "for country,disset in water_country_selected_dis_dict.items():\n",
    "    dis_dict = {}\n",
    "    for dis in disset:\n",
    "        z_list = [] \n",
    "        if len(water_country_selected_chem_id_dict[country])>0:\n",
    "            for chem in water_country_selected_chem_id_dict[country]:\n",
    "                try:\n",
    "                    z_list.append(real_significant_final_exp_proximity[chem,dis][1])\n",
    "                except:\n",
    "                    pass\n",
    "        dis_dict[dis]= np.round(np.mean(z_list),2)\n",
    "                    \n",
    "    country_water_dis_mean_prox_dict[country] = dis_dict"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "6e7d95a0-7390-40cf-91bd-37acbc535ce3",
   "metadata": {},
   "outputs": [],
   "source": [
    "country_water_chem_id_gbd_dict = {}\n",
    "\n",
    "for country,dis_dict in country_water_dis_mean_prox_dict.items():\n",
    "    if len(dis_dict)>0:\n",
    "        dis_z_list = []\n",
    "        for dis,z in dis_dict.items():\n",
    "            try:\n",
    "                dis_z_list.append([water_disgenet_gbd_dict[dis],z])\n",
    "            except:\n",
    "                pass\n",
    "        country_water_chem_id_gbd_dict[country] = dis_z_list"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "8915ca4f-7b0c-4662-b57c-59eb56724f04",
   "metadata": {},
   "outputs": [],
   "source": [
    "country_water_chem_id_gbd_mean_dict = {}\n",
    "for country,dislist in country_water_chem_id_gbd_dict.items():\n",
    "    dis_dict = {}\n",
    "    dis_list = []\n",
    "    for el in dislist:\n",
    "        dis_list.append(el[0]) \n",
    "    dis_list_counter_dict = dict(Counter(dis_list).most_common())\n",
    "    for dis,count in dis_list_counter_dict.items():\n",
    "        z_list = []\n",
    "        if count>1:\n",
    "            for el in dislist:\n",
    "                if dis==el[0]:\n",
    "                    z_list.append(el[1])\n",
    "            dis_dict[dis]=np.round(np.mean(z_list),2)\n",
    "        else:\n",
    "            for el in dislist:\n",
    "                if dis==el[0]:\n",
    "                    dis_dict[el[0]] = el[1]\n",
    "    country_water_chem_id_gbd_mean_dict[country]=dis_dict\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "24d03627-3243-4242-aa67-12f4cbe8f383",
   "metadata": {},
   "outputs": [],
   "source": [
    "country_water_faraway_dis_dict={}\n",
    "for country,chemset in water_country_selected_chem_id_dict.items():\n",
    "    if len(chemset)>0:\n",
    "        far_away_dis_set = []\n",
    "        for chem in chemset:\n",
    "            for chem_dis_pair in faraway_exp_dis_dict.keys():\n",
    "                if chem in chem_dis_pair[0]:\n",
    "                    far_away_dis_set.append(chem_dis_pair[1])\n",
    "        country_water_faraway_dis_dict[country] = set(far_away_dis_set)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "6e45a54e-5743-4382-9c5d-af1f584ce501",
   "metadata": {},
   "outputs": [],
   "source": [
    "faraway_water_disset = []\n",
    "\n",
    "for country,disset in country_water_faraway_dis_dict.items():\n",
    "    for dis in disset:\n",
    "        if dis not in faraway_water_disset:\n",
    "            faraway_water_disset.append(dis)\n",
    "\n",
    "print(len(faraway_water_disset))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "531eb327-4227-4e95-9e33-a2c7067bd421",
   "metadata": {},
   "outputs": [],
   "source": [
    "n_long_faraway_water_disset_gbd_dict = {}\n",
    "for dis1 in faraway_water_disset:\n",
    "    dis1_processed = preprocess_string(dis1)\n",
    "    # Initialize best match tracking\n",
    "    best_match = (None, 0)  # (Best match disease name, similarity score)\n",
    "    for dis2 in disease_set:\n",
    "        dis2_processed = preprocess_string(dis2)\n",
    "        similarity = get_similarity_score(dis1_processed, dis2_processed)\n",
    "        # Update if this is the best match so far\n",
    "        if similarity > best_match[1]:\n",
    "            best_match = (dis2, similarity)\n",
    "    \n",
    "    # You may want to set a threshold for \"good enough\" matching\n",
    "    if best_match[1] > 50:  # This threshold\n",
    "        n_long_faraway_water_disset_gbd_dict[dis1] = best_match[0]\n",
    "\n",
    "print(len(n_long_faraway_water_disset_gbd_dict))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "7a89f5a9-bd5f-440d-bafa-d8e32b975329",
   "metadata": {},
   "outputs": [],
   "source": [
    "keys_to_delete=['Weill-Marchesani-Like Syndrome','HYPOGONADOTROPIC HYPOGONADISM 11 WITH OR WITHOUT ANOSMIA','Night blindness, congenital stationary',\n",
    "             'Narcolepsy 1', 'DEAFNESS, X-LINKED 4 (disorder)','Juvenile nephropathic cystinosis (disorder)', 'Congenital Pain Insensitivity',\n",
    "                'Nephropathic cystinosis','Congenital long QT syndrome','Glomerulocystic Kidney Disease with Hyperuricemia and Isosthenuria',\n",
    "                'Keutel syndrome','FICOLIN 3 DEFICIENCY','Autosomal Dominant Nocturnal Frontal Lobe Epilepsy','Cavitary Optic Disc Anomalies',\n",
    "                'alpha, alpha-Trehalase deficiency','Nephritis, Tubulointerstitial','Cerebral ventriculomegaly','HYPOGONADOTROPIC HYPOGONADISM 4 WITH OR WITHOUT ANOSMIA',\n",
    "                'Goiter','Medullary Cystic Kidney Disease Type 2','Malabsorption Syndrome','CATARACT 39, MULTIPLE TYPES',\n",
    "                'Long Qt Syndrome 10','RETINITIS PIGMENTOSA 49','Olmsted syndrome','Cystinosis','HIRSCHSPRUNG DISEASE, SUSCEPTIBILITY TO, 4',\n",
    "                'MASP2 Deficiency','Cone monochromatism','Enterokinase Deficiency','Retinitis Pigmentosa 7','Narcolepsy','Jarcho-Levin syndrome',\n",
    "                'Waardenburg Syndrome, Type 4b','NIGHT BLINDNESS, CONGENITAL STATIONARY, TYPE 1E','Factor XIII, B Subunit, Deficiency Of',\n",
    "                'Renal hypouricemia','MYASTHENIC SYNDROME, CONGENITAL, 1A, SLOW-CHANNEL','BLEEDING DISORDER, PLATELET-TYPE, 20',\n",
    "                'SMALL PATELLA SYNDROME','Amish Infantile Epilepsy Syndrome','Scoliosis, unspecified','Focal glomerulosclerosis',\n",
    "                'Meningococcal Infections','Indifference to Pain, Congenital, Autosomal Recessive','Neuropathy, Distal Hereditary Motor, Type VIIA',\n",
    "                'Obstruction of pelviureteric junction','TELANGIECTASIA, HEREDITARY HEMORRHAGIC, TYPE 5','Marshall syndrome','STICKLER SYNDROME, TYPE II (disorder)',\n",
    "                'Histiocytosis','Hypopigmentation disorder','PAROXYSMAL EXTREME PAIN DISORDER','NEMALINE MYOPATHY 10','DEAFNESS, AUTOSOMAL RECESSIVE 42 (disorder)',\n",
    "                'Nivelon Nivelon Mabille syndrome','HERMANSKY-PUDLAK SYNDROME 4','Juvenile hemochromatosis','HEMOCHROMATOSIS, TYPE 2A','USHER SYNDROME, TYPE IJ',\n",
    "                'CATARACT 47','Usher Syndrome, Type IF','Hypouricemia, Renal, 2','NIGHT BLINDNESS, CONGENITAL STATIONARY, TYPE 1G','Primary Erythermalgia',\n",
    "                'Trimethylaminuria','Hypertrichosis','Congenital disorder of glycosylation, type 2C','Night Blindness, Congenital Stationary, Autosomal Dominant 3',\n",
    "                'Lysinuric Protein Intolerance','RETINITIS PIGMENTOSA, JUVENILE, LRAT-RELATED','Trichotillomania','Hydrocephalus','Usher Syndrome, Type I',\n",
    "                'STARGARDT DISEASE 1 (disorder)','Perisylvian syndrome','Liver Failure','CATARACT 4, MULTIPLE TYPES','DEAFNESS, AUTOSOMAL RECESSIVE 1A (disorder)',\n",
    "                'CILIARY DYSKINESIA, PRIMARY, 7 (disorder)','Vesico-Ureteral Reflux','VAN DER WOUDE SYNDROME 1','Hyperaldosteronism','Popliteal pterygium syndrome',\n",
    "                'Riboflavin Deficiency','Zunich neuroectodermal syndrome','Hypercalcemia, Infantile','CONE-ROD DYSTROPHY 3 (disorder)','RETINITIS PIGMENTOSA 19','CONGENITAL DISORDER OF GLYCOSYLATION, TYPE Ip',\n",
    "                'Juvenile macular degeneration',\"Stargardt's disease\",'RETINITIS PIGMENTOSA 48'\n",
    "                \n",
    "]\n",
    "\n",
    "n_long_faraway_water_disset_gbd_dict_cleaned_v1 = {}\n",
    "for disgenet_dis, gbd_dis in n_long_faraway_water_disset_gbd_dict.items():\n",
    "    if disgenet_dis not in keys_to_delete:\n",
    "        n_long_faraway_water_disset_gbd_dict_cleaned_v1[disgenet_dis] = gbd_dis\n",
    "        "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d269d82c-e74e-42d7-9121-11c480e3d04c",
   "metadata": {},
   "outputs": [],
   "source": [
    "keys_to_delete_v2 = ['JOUBERT SYNDROME 16','Leber Congenital Amaurosis 14','Dursun Syndrome','Neutropenia, Severe Congenital, Autosomal Recessive 4',\n",
    "                     'Dihydropyrimidinase deficiency','Deuteranomaly','CRANIOSYNOSTOSIS AND DENTAL ANOMALIES','Canavan Disease','Canavan Disease, Juvenile',\n",
    "                     'Dimethylglycine Dehydrogenase Deficiency','Retinitis Pigmentosa','Amnesia','POLYARTERITIS NODOSA, CHILDHOOD-ONSET','RETINITIS PIGMENTOSA 30',\n",
    "                     'NIGHT BLINDNESS, CONGENITAL STATIONARY, TYPE 1B','Weill-Marchesani syndrome','Ectodermal Dysplasia','MACULAR DYSTROPHY, PATTERNED, 1',\n",
    "                     'STEEL SYNDROME','CONGENITAL DISORDER OF GLYCOSYLATION, TYPE Iu','Pfaundler-Hurler Syndrome','CATARACT 19, MULTIPLE TYPES',\n",
    "                     'RETINITIS PIGMENTOSA 45','HYPERGLYCINURIA (disorder)','Arterial calcification of infancy','HYPOGONADOTROPIC HYPOGONADISM 3 WITH OR WITHOUT ANOSMIA',\n",
    "                     'Muscle eye brain disease','Epilepsy','SEIZURES, BENIGN FAMILIAL INFANTILE, 3','Deafness, Autosomal Dominant 3B','West Syndrome',\n",
    "                     'Multiple synostoses syndrome 1','Hurler-Scheie Syndrome','Bardet-Biedl Syndrome','Sneddon Syndrome','Idiopathic Pulmonary Fibrosis',\n",
    "                     'Holoprosencephaly','MYASTHENIC SYNDROME, CONGENITAL, 4A, SLOW-CHANNEL','Leber Congenital Amaurosis','CEROID LIPOFUSCINOSIS, NEURONAL, 13',\n",
    "                     'Hermanski-Pudlak Syndrome','Pain','Gilles de la Tourette syndrome','Canavan Disease, Infantile','Congenital hemihypertrophy',\n",
    "                     'Myasthenic Syndromes, Congenital','Meleda Disease','Intervertebral disc disorder','Nonsyndromic Deafness'\n",
    "    \n",
    "]\n",
    "\n",
    "\n",
    "\n",
    "n_long_faraway_water_disset_gbd_dict_cleaned_v2 = {}\n",
    "for disgenet_dis, gbd_dis in n_long_faraway_water_disset_gbd_dict_cleaned_v1.items():\n",
    "    if disgenet_dis not in keys_to_delete_v2:\n",
    "        n_long_faraway_water_disset_gbd_dict_cleaned_v2[disgenet_dis] = gbd_dis\n",
    "\n",
    "n_long_faraway_water_disset_gbd_dict_cleaned_v2['Epilepsy'] = 'Idiopathic epilepsy'\n",
    "n_long_faraway_water_disset_gbd_dict_cleaned_v2['SEIZURES, BENIGN FAMILIAL INFANTILE, 3'] = 'Idiopathic epilepsy'\n",
    "n_long_faraway_water_disset_gbd_dict_cleaned_v2['CATARACT 39, MULTIPLE TYPES'] = 'Cataract'\n",
    "n_long_faraway_water_disset_gbd_dict_cleaned_v2['NIGHT BLINDNESS, CONGENITAL STATIONARY, TYPE 1E'] = 'Blindness and vision loss'\n",
    "n_long_faraway_water_disset_gbd_dict_cleaned_v2['CATARACT 4, MULTIPLE TYPES'] = 'Cataract'\n",
    "n_long_faraway_water_disset_gbd_dict_cleaned_v2['CATARACT 47'] = 'Cataract'\n",
    "\n",
    "\n",
    "water_faraway_dis_list = list(n_long_faraway_water_disset_gbd_dict_cleaned_v2.values())\n",
    "water_faraway_dis_list.extend(['Congenital musculoskeletal and limb anomalies','Congenital birth defects','Congenital heart anomalies'])\n",
    "\n",
    "print(len(water_faraway_dis_list))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "6cea5aff-a4fb-44d9-bb8d-edb8cf410d99",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "faraway_manually_added_diseases = ['Congenital musculoskeletal and limb anomalies','Congenital birth defects','Congenital heart anomalies']\n",
    "\n",
    "country_water_faraway_dis_dict = {}\n",
    "for country,dis_dict in country_water_chem_id_gbd_mean_dict.items():\n",
    "    chem_set = water_country_selected_chem_id_dict[country]\n",
    "    country_faraway_dis_list = []\n",
    "    country_specific_faraway_dis_dict = {}\n",
    "    \n",
    "    for dis in water_faraway_dis_list:\n",
    "        if dis not in dis_dict.keys():\n",
    "            country_faraway_dis_list.append(dis)\n",
    "    for dis in country_faraway_dis_list:\n",
    "        if dis not in faraway_manually_added_diseases:\n",
    "            disgenet_name_list = []\n",
    "            z_score_list = []\n",
    "            for disgenet_name,gbd_id in n_long_faraway_water_disset_gbd_dict_cleaned_v2.items():\n",
    "                if dis == gbd_id:\n",
    "                    disgenet_name_list.append(disgenet_name)\n",
    "                    for chem in chem_set:\n",
    "                        try:\n",
    "                            z_score_list.append(exp_proximity_not_significant[chem,disgenet_name][1])\n",
    "                        except:\n",
    "                            pass\n",
    "            if np.mean(z_score_list)>1.5:\n",
    "                country_specific_faraway_dis_dict[dis]=np.mean(z_score_list)\n",
    "        elif dis=='Congenital heart anomalies':\n",
    "            disgenet_name_list = ['Congenital Heart Defects','Deafness, Congenital Heart Defects, and Posterior Embryotoxon',\n",
    "                                 'Congenital long QT syndrome']\n",
    "            z_score_list = []\n",
    "            for disgenet_name in disgenet_name_list:\n",
    "                for chem in chem_set:\n",
    "                    try:\n",
    "                        z_score_list.append(exp_proximity_not_significant[chem,disgenet_name][1])\n",
    "                    except:\n",
    "                        pass\n",
    "            if np.mean(z_score_list)>1.5:\n",
    "                country_specific_faraway_dis_dict[dis]=np.mean(z_score_list)\n",
    "        \n",
    "        elif dis=='Congenital musculoskeletal and limb anomalies':\n",
    "            disgenet_name_list = ['Congenital Fiber Type Disproportion','Myopathy, Actin, Congenital, with Excess of Thin Myofilaments',\n",
    "                                 'Myopathy, Actin, Congenital, With Cores','Muscular Dystrophy, Congenital, Megaconial Type',\n",
    "                                 'Myotonia Congenita','Myasthenic Syndromes, Congenital','Muscular Dystrophy, Congenital, Megaconial Type',\n",
    "                                 'Congenital Myasthenic Syndromes, Postsynaptic','Congenital Myasthenic Syndromes, Presynaptic',\n",
    "                                 'Muscular Dystrophy, Congenital, Due To Integrin Alpha-7 Deficiency','Muscular Dystrophy, Congenital, Type 1D',\n",
    "                                 'Congenital Fibrosis of the Extraocular Muscles']\n",
    "            z_score_list = []\n",
    "            for disgenet_name in disgenet_name_list:\n",
    "                for chem in chem_set:\n",
    "                    try:\n",
    "                        z_score_list.append(exp_proximity_not_significant[chem,disgenet_name][1])\n",
    "                    except:\n",
    "                        pass\n",
    "            if np.mean(z_score_list)>1.5:\n",
    "                country_specific_faraway_dis_dict[dis]=np.mean(z_score_list)\n",
    "            \n",
    "        elif dis=='Congenital birth defects':\n",
    "            disgenet_name_list = [dis for dis in list(diseases_genes_associated_ppi.keys()) if 'Congenital' in dis or 'congenital' in dis]\n",
    "            z_score_list = []\n",
    "            for disgenet_name in disgenet_name_list:\n",
    "                for chem in chem_set:\n",
    "                    try:\n",
    "                        z_score_list.append(exp_proximity_not_significant[chem,disgenet_name][1])\n",
    "                    except:\n",
    "                        pass\n",
    "            if np.mean(z_score_list)>1.5:\n",
    "                country_specific_faraway_dis_dict[dis]=np.mean(z_score_list)\n",
    "\n",
    "    country_water_faraway_dis_dict[country] = country_specific_faraway_dis_dict\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "f187c422-0a42-41c3-9f52-e44302f20aec",
   "metadata": {},
   "outputs": [],
   "source": [
    "#We will use the same no-proximal disease sets/dictionary that we have calcualted for the air pollution\n",
    "\n",
    "country_noprox_water_dis_dict = {}\n",
    "for country,dis_dict in country_water_chem_id_gbd_mean_dict.items():\n",
    "    n_dis = len(dis_dict)\n",
    "    chem_set = water_country_selected_chem_id_dict[country]\n",
    "    country_noprox_dis_list = []\n",
    "    country_specific_noprox_dis_dict = {}\n",
    "    \n",
    "    for gbd_dis,dislist in no_prox_air_disset_gbd_dict_cleaned_inv.items():\n",
    "        if gbd_dis not in dis_dict.keys():\n",
    "            if gbd_dis not in list(country_water_faraway_dis_dict[country].keys()):\n",
    "                country_noprox_dis_list.append(gbd_dis)\n",
    "    for dis in country_noprox_dis_list:\n",
    "        disgenet_name_list = no_prox_air_disset_gbd_dict_cleaned_inv[dis]\n",
    "        z_score_list = []\n",
    "        for chem in chem_set:\n",
    "            for disgenet_name in disgenet_name_list:\n",
    "                try:\n",
    "                    z_score_list.append(exp_proximity_not_significant[chem,disgenet_name][1])\n",
    "                except:\n",
    "                    pass\n",
    "        if np.mean(z_score_list)>-1.5:\n",
    "            country_specific_noprox_dis_dict[dis]=np.mean(z_score_list)\n",
    "    country_specific_noprox_dis_dict_sort = {k: v for k, v in sorted(country_specific_noprox_dis_dict.items(), key=lambda item: item[1])}\n",
    "    \n",
    "    if len(country_specific_noprox_dis_dict_sort)>n_dis:\n",
    "        country_specific_noprox_dis_dict_sort_keys = list(country_specific_noprox_dis_dict_sort.keys())\n",
    "        n_country_specific_noprox_dis_dict_sort = {}\n",
    "        for k in list(country_specific_noprox_dis_dict_sort_keys)[:n_dis]:\n",
    "            n_country_specific_noprox_dis_dict_sort[k]=country_specific_noprox_dis_dict_sort[k]\n",
    "        \n",
    "        country_noprox_water_dis_dict[country] = n_country_specific_noprox_dis_dict_sort\n",
    "    else:\n",
    "        country_noprox_water_dis_dict[country] = country_specific_noprox_dis_dict_sort\n",
    " "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "cd7f8b3e-9f0b-420c-82af-48fd72aae8e5",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Let's just merge the no-proximal dictionary and the faraway dictionary\n",
    "new_water_country_distant_dis_dict = {}\n",
    "for country,dis_dict in country_water_chem_id_gbd_mean_dict.items():\n",
    "    n_proxdis = len(dis_dict)\n",
    "    n_distantdis = len(country_water_faraway_dis_dict[country])\n",
    "    noprox_list = list(country_noprox_water_dis_dict[country].keys())\n",
    "    missing_dis = n_proxdis-n_distantdis\n",
    "    try:\n",
    "        dislist = noprox_list[:missing_dis]\n",
    "        distant_dict = country_water_faraway_dis_dict[country].copy()\n",
    "        for dis in dislist:\n",
    "            distant_dict[dis] = country_noprox_water_dis_dict[country][dis]\n",
    "        new_water_country_distant_dis_dict[country] = distant_dict\n",
    "    except:\n",
    "        distant_dict = {**country_water_faraway_dis_dict[country], **country_noprox_water_dis_dict[country]}\n",
    "        new_water_country_distant_dis_dict[country] = distant_dict"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "da8965a3-cf0b-42ae-a72c-c10d818f70b6",
   "metadata": {},
   "outputs": [],
   "source": [
    "proximal_dis_list=[] \n",
    "for country, disdict in country_water_chem_id_gbd_mean_dict.items():\n",
    "    proximal_dis_list.extend(list(disdict.keys()))\n",
    "    \n",
    "proximal_dis_incidence={}\n",
    "country_set=set(GBD_df['location'])\n",
    "for dis in set(proximal_dis_list):\n",
    "    incidence_list=[]\n",
    "    for country in country_set:\n",
    "        incidence_list.append(country_dis_incidence[country][dis])\n",
    "    proximal_dis_incidence[dis]=incidence_list\n",
    "\n",
    "proximal_dis_prevalence={}\n",
    "for dis in set(proximal_dis_list):\n",
    "    prevalence_list=[]\n",
    "    for country in country_set:\n",
    "        prevalence_list.append(country_dis_prevalence[country][dis])\n",
    "    proximal_dis_prevalence[dis]=prevalence_list\n",
    "\n",
    "proximal_dis_mortality={}\n",
    "for dis in set(proximal_dis_list):\n",
    "    mortality_list=[]\n",
    "    for country in country_set:\n",
    "        mortality_list.append(country_dis_mortality[country][dis])\n",
    "    proximal_dis_mortality[dis]=mortality_list\n",
    "\n",
    "proximal_dis_daly={}\n",
    "for dis in set(proximal_dis_list):\n",
    "    daly_list=[]\n",
    "    for country in country_set:\n",
    "        daly_list.append(country_dis_DALYs[country][dis])\n",
    "    proximal_dis_daly[dis]=daly_list\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "f70feecd-f4c5-486c-b75e-03ad1df494b6",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "water_country_proximal_zscore_dict={}\n",
    "for country,dis_dict in country_water_chem_id_gbd_mean_dict.items():\n",
    "    country_proximal_zscore_list=[]\n",
    "    for dis in list(dis_dict.keys()):\n",
    "        mu = np.mean(proximal_dis_incidence[dis])\n",
    "        std = np.std(proximal_dis_incidence[dis])\n",
    "\n",
    "        z = (country_dis_incidence[country][dis]-mu)/std\n",
    "        country_proximal_zscore_list.append(z)\n",
    "\n",
    "        from scipy.stats import norm\n",
    "        ptemp = norm.sf(abs(z))\n",
    "    water_country_proximal_zscore_dict[country]=country_proximal_zscore_list\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "6889f529-32af-4ff0-b43f-118d0d3b9233",
   "metadata": {},
   "outputs": [],
   "source": [
    "distant_dis_list=[] \n",
    "for country, disdict in new_water_country_distant_dis_dict.items():\n",
    "    distant_dis_list.extend(list(disdict.keys()))\n",
    "    \n",
    "distant_dis_incidence={}\n",
    "country_set=set(GBD_df['location'])\n",
    "for dis in set(distant_dis_list):\n",
    "    incidence_list=[]\n",
    "    for country in country_set:\n",
    "        incidence_list.append(country_dis_incidence[country][dis])\n",
    "    distant_dis_incidence[dis]=incidence_list\n",
    "\n",
    "\n",
    "distant_dis_prevalence={}\n",
    "for dis in set(distant_dis_list):\n",
    "     prevalence_list=[]\n",
    "     for country in country_set:\n",
    "         prevalence_list.append(country_dis_prevalence[country][dis])\n",
    "     distant_dis_prevalence[dis]=prevalence_list\n",
    "\n",
    "distant_dis_mortality={}\n",
    "for dis in set(distant_dis_list):\n",
    "     mortality_list=[]\n",
    "     for country in country_set:\n",
    "         mortality_list.append(country_dis_mortality[country][dis])\n",
    "     distant_dis_mortality[dis]=mortality_list\n",
    "\n",
    "distant_dis_daly={}\n",
    "for dis in set(distant_dis_list):\n",
    "     daly_list=[]\n",
    "     for country in country_set:\n",
    "         daly_list.append(country_dis_DALYs[country][dis])\n",
    "     distant_dis_daly[dis]=daly_list\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "e2018ff9-eddc-4a62-aac7-cf2576cd6bf0",
   "metadata": {},
   "outputs": [],
   "source": [
    "water_country_distant_zscore_dict={}\n",
    "\n",
    "for country,dis_dict in new_water_country_distant_dis_dict.items():\n",
    "    country_distant_zscore_list=[]\n",
    "    for dis in list(dis_dict.keys()):\n",
    "        mu = np.mean(distant_dis_incidence[dis])\n",
    "        std = np.std(distant_dis_incidence[dis])\n",
    "\n",
    "        z = (country_dis_incidence[country][dis]-mu)/std\n",
    "        country_distant_zscore_list.append(z)\n",
    "\n",
    "        from scipy.stats import norm\n",
    "        ptemp = norm.sf(abs(z))\n",
    "    water_country_distant_zscore_dict[country]=country_distant_zscore_list\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "e956dde9-05d9-4f8a-a656-7b2ede17cf3a",
   "metadata": {},
   "outputs": [],
   "source": [
    "water_pollution_zscore_df = pd.DataFrame()\n",
    "\n",
    "selected_countries=set(water_country_distant_zscore_dict.keys())\n",
    "tot_country_list=[]\n",
    "tot_zscore_list=[]\n",
    "location_network_list=[]\n",
    "for country in selected_countries:\n",
    "    farawaylist_cleaned = [x for x in water_country_distant_zscore_dict[country] if str(x) != 'nan']\n",
    "    proximallist_cleaned = [x for x in water_country_proximal_zscore_dict[country] if str(x) != 'nan']\n",
    "    tot_zscorelist_partial=farawaylist_cleaned+proximallist_cleaned\n",
    "    for i in proximallist_cleaned:\n",
    "        location_network_list.append(\"proximal\")\n",
    "        tot_zscore_list.append(i)\n",
    "        tot_country_list.append(country)\n",
    "    for i in farawaylist_cleaned:\n",
    "        location_network_list.append(\"distant\")\n",
    "        tot_zscore_list.append(i)\n",
    "        tot_country_list.append(country)\n",
    "\n",
    "water_pollution_zscore_df['location']=tot_country_list\n",
    "water_pollution_zscore_df['z-score']=tot_zscore_list\n",
    "water_pollution_zscore_df['network_location']=location_network_list"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "faad6e73-8bf3-4880-85e4-03b4633662d7",
   "metadata": {},
   "outputs": [],
   "source": [
    "my_order = sorted(list(set(water_pollution_zscore_df['location'])))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "7156cf9b-4806-4645-987f-b3dbe3609401",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Let's test the difference between proximal and no-proximal (because with distant is kind of obvious and it doesn't\n",
    "#make so much sense)\n",
    "import scipy\n",
    "for country in water_country_proximal_zscore_dict.keys():\n",
    "    try:\n",
    "        proximallist_cleaned = [x for x in water_country_proximal_zscore_dict[country] if str(x) != 'nan']\n",
    "        distantlist_cleaned = [x for x in water_country_distant_zscore_dict[country] if str(x) != 'nan']\n",
    "        prox_mean = np.mean(proximallist_cleaned)\n",
    "        distant_mean = np.mean(distantlist_cleaned)\n",
    "        pval_noprox = scipy.stats.mannwhitneyu(proximallist_cleaned, distantlist_cleaned, use_continuity=True, alternative='greater')[1]\n",
    "        print(country,prox_mean,distant_mean,round(pval_noprox,4),round(pval_distant,4))\n",
    "    except:\n",
    "        print(country,prox_mean,distant_mean,round(pval_noprox,4))\n",
    "        "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "7b817fdc-21dd-4a2d-a148-4ac0e29aaca6",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "fig, ax = plt.subplots(figsize=(15, 20))\n",
    "sns.set(style=\"white\")\n",
    "g = sns.boxplot(x=\"z-score\", y=\"location\", orient=\"h\",\n",
    "                hue=\"network_location\", palette=[\"r\", \"b\"], order=my_order,\n",
    "                data=water_pollution_zscore_df)\n",
    "fontsize = 36\n",
    "plt.legend(loc=\"upper right\", frameon=True, fontsize=fontsize)\n",
    "plt.yticks(fontsize=36)\n",
    "\n",
    "for ax in plt.gcf().axes:\n",
    "    l = ax.get_xlabel()\n",
    "    ax.set_xlabel(l, fontsize=36)\n",
    "    ax.set_ylabel(l, fontsize=36)\n",
    "\n",
    "plt.ylabel('', fontsize=25)\n",
    "plt.xlabel('', fontsize=25)\n",
    "ax.spines['top'].set_visible(False)\n",
    "ax.spines['right'].set_visible(False)\n",
    "ax.spines['bottom'].set_visible(True)\n",
    "ax.spines['left'].set_visible(True)\n",
    "ax.set_xticklabels(ax.get_xticks(), size=36)\n",
    "\n",
    "pv_list = []  # Example p-values\n",
    "locations = []  # Example y-axis categories\n",
    "stars = []  # Based on the p-values, this is what we want to add\n",
    "\n",
    "\n",
    "\n",
    "for country in my_order:\n",
    "    proximallist_cleaned = [x for x in water_country_proximal_zscore_dict[country] if str(x) != 'nan']\n",
    "    distantlist_cleaned = [x for x in water_country_distant_zscore_dict[country] if str(x) != 'nan']\n",
    "    prox_mean = np.mean(proximallist_cleaned)\n",
    "    distant_mean = np.mean(distantlist_cleaned)\n",
    "    pval_noprox = scipy.stats.mannwhitneyu(proximallist_cleaned, distantlist_cleaned, use_continuity=True, alternative='greater')[1]\n",
    "    pv_list.append(pval_noprox)\n",
    "    locations.append(country)\n",
    "    if pval_noprox<0.001:\n",
    "        stars.append(\"***\")\n",
    "    elif pval_noprox>0.001 and pval_noprox<0.01:\n",
    "        stars.append(\"**\")\n",
    "    elif pval_noprox>0.01 and pval_noprox<0.05:\n",
    "        stars.append(\"*\")\n",
    "    else:\n",
    "        stars.append(\"n.s.\")\n",
    "\n",
    "\n",
    "for pv, loc, star in zip(pv_list, locations, stars):\n",
    "    yloc = my_order.index(loc) + 0.3  # This presumes 'my_order' list contains the order of y-categories\n",
    "    xloc = max(water_pollution_zscore_df[water_pollution_zscore_df['location'] == loc]['z-score']) + 0.5  \n",
    "    # Add the annotation\n",
    "    ax.text(xloc, yloc, star, fontsize=fontsize, color=\"black\", ha='center')\n",
    "\n",
    "#plt.savefig('output/Figures/Figure4F_v09.pdf',format='pdf',dpi=300, bbox_inches = \"tight\")\n",
    "# Once everything is set, display or save the figure\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "50750b96-1853-441e-9107-20b8315666d7",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c36e044b-c018-4b13-9123-fea142054b3d",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Prevalence\n",
    "\n",
    "water_country_proximal_zscore_dict={}\n",
    "for country,dis_dict in country_water_chem_id_gbd_mean_dict.items():\n",
    "    country_proximal_zscore_list=[]\n",
    "    for dis in list(dis_dict.keys()):\n",
    "        mu = np.mean(proximal_dis_prevalence[dis])\n",
    "        std = np.std(proximal_dis_prevalence[dis])\n",
    "        z = (country_dis_prevalence[country][dis]-mu)/std\n",
    "        country_proximal_zscore_list.append(z)\n",
    "        from scipy.stats import norm\n",
    "        ptemp = norm.sf(abs(z))\n",
    "    water_country_proximal_zscore_dict[country]=country_proximal_zscore_list\n",
    "\n",
    "\n",
    "water_country_distant_zscore_dict={}\n",
    "for country,dis_dict in new_water_country_distant_dis_dict.items():\n",
    "    country_distant_zscore_list=[]\n",
    "    for dis in list(dis_dict.keys()):\n",
    "        mu = np.mean(distant_dis_prevalence[dis])\n",
    "        std = np.std(distant_dis_prevalence[dis])\n",
    "        z = (country_dis_prevalence[country][dis]-mu)/std\n",
    "        country_distant_zscore_list.append(z)\n",
    "        from scipy.stats import norm\n",
    "        ptemp = norm.sf(abs(z))\n",
    "    water_country_distant_zscore_dict[country]=country_distant_zscore_list\n",
    "\n",
    "water_pollution_zscore_df = pd.DataFrame()\n",
    "selected_countries=set(water_country_distant_zscore_dict.keys())\n",
    "tot_country_list=[]\n",
    "tot_zscore_list=[]\n",
    "location_network_list=[]\n",
    "for country in selected_countries:\n",
    "    farawaylist_cleaned = [x for x in water_country_distant_zscore_dict[country] if str(x) != 'nan']\n",
    "    proximallist_cleaned = [x for x in water_country_proximal_zscore_dict[country] if str(x) != 'nan']\n",
    "    tot_zscorelist_partial=farawaylist_cleaned+proximallist_cleaned\n",
    "    for i in proximallist_cleaned:\n",
    "        location_network_list.append(\"proximal\")\n",
    "        tot_zscore_list.append(i)\n",
    "        tot_country_list.append(country)\n",
    "    for i in farawaylist_cleaned:\n",
    "        location_network_list.append(\"distant\")\n",
    "        tot_zscore_list.append(i)\n",
    "        tot_country_list.append(country)\n",
    "\n",
    "water_pollution_zscore_df['location']=tot_country_list\n",
    "water_pollution_zscore_df['z-score']=tot_zscore_list\n",
    "water_pollution_zscore_df['network_location']=location_network_list\n",
    "\n",
    "for country in water_country_proximal_zscore_dict.keys():\n",
    "    try:\n",
    "        proximallist_cleaned = [x for x in water_country_proximal_zscore_dict[country] if str(x) != 'nan']\n",
    "        distantlist_cleaned = [x for x in water_country_distant_zscore_dict[country] if str(x) != 'nan']\n",
    "        prox_mean = np.mean(proximallist_cleaned)\n",
    "        distant_mean = np.mean(distantlist_cleaned)\n",
    "        pval_noprox = scipy.stats.mannwhitneyu(proximallist_cleaned, distantlist_cleaned, use_continuity=True, alternative='greater')[1]\n",
    "        print(country,prox_mean,distant_mean,round(pval_noprox,4),round(pval_distant,4))\n",
    "    except:\n",
    "        print(country,prox_mean,distant_mean,round(pval_noprox,4))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "0738cb7c-c701-4f3f-b75f-a40111d5fb37",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "fig, ax = plt.subplots(figsize=(15, 20))\n",
    "sns.set(style=\"white\")\n",
    "g = sns.boxplot(x=\"z-score\", y=\"location\", orient=\"h\",\n",
    "                hue=\"network_location\", palette=[\"#2B7A98\", \"#54CAF2\"], order=my_order,\n",
    "                data=water_pollution_zscore_df)\n",
    "fontsize = 36\n",
    "plt.legend(loc=\"upper right\", frameon=True, fontsize=fontsize)\n",
    "plt.yticks(fontsize=36)\n",
    "\n",
    "for ax in plt.gcf().axes:\n",
    "    l = ax.get_xlabel()\n",
    "    ax.set_xlabel(l, fontsize=36)\n",
    "    ax.set_ylabel(l, fontsize=36)\n",
    "\n",
    "plt.ylabel('', fontsize=25)\n",
    "plt.xlabel('', fontsize=25)\n",
    "ax.spines['top'].set_visible(False)\n",
    "ax.spines['right'].set_visible(False)\n",
    "ax.spines['bottom'].set_visible(True)\n",
    "ax.spines['left'].set_visible(True)\n",
    "ax.set_xticklabels(ax.get_xticks(), size=36)\n",
    "\n",
    "pv_list = []  # Example p-values\n",
    "locations = []  # Example y-axis categories\n",
    "stars = []  # Based on the p-values, this is what we want to add\n",
    "\n",
    "\n",
    "\n",
    "for country in my_order:\n",
    "    proximallist_cleaned = [x for x in water_country_proximal_zscore_dict[country] if str(x) != 'nan']\n",
    "    distantlist_cleaned = [x for x in water_country_distant_zscore_dict[country] if str(x) != 'nan']\n",
    "    prox_mean = np.mean(proximallist_cleaned)\n",
    "    distant_mean = np.mean(distantlist_cleaned)\n",
    "    pval_noprox = scipy.stats.mannwhitneyu(proximallist_cleaned, distantlist_cleaned, use_continuity=True, alternative='greater')[1]\n",
    "    pv_list.append(pval_noprox)\n",
    "    locations.append(country)\n",
    "    if pval_noprox<0.001:\n",
    "        stars.append(\"***\")\n",
    "    elif pval_noprox>0.001 and pval_noprox<0.01:\n",
    "        stars.append(\"**\")\n",
    "    elif pval_noprox>0.01 and pval_noprox<0.05:\n",
    "        stars.append(\"*\")\n",
    "    else:\n",
    "        stars.append(\"n.s.\")\n",
    "\n",
    "\n",
    "for pv, loc, star in zip(pv_list, locations, stars):\n",
    "    yloc = my_order.index(loc) + 0.3  # This presumes 'my_order' list contains the order of y-categories\n",
    "    xloc = max(water_pollution_zscore_df[water_pollution_zscore_df['location'] == loc]['z-score']) + 0.5  \n",
    "    # Add the annotation\n",
    "    ax.text(xloc, yloc, star, fontsize=fontsize, color=\"black\", ha='center')\n",
    "\n",
    "plt.savefig('output/Figures/SD12D_v01_post_rev.pdf',format='pdf',dpi=300, bbox_inches = \"tight\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ea0c2cd4-29ea-4136-a1b5-920ad59cdbe7",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "86692bf1-c0b0-4a6b-aa30-22aeb8ca1438",
   "metadata": {},
   "outputs": [],
   "source": [
    "#mortality\n",
    "\n",
    "water_country_proximal_zscore_dict={}\n",
    "for country,dis_dict in country_water_chem_id_gbd_mean_dict.items():\n",
    "    country_proximal_zscore_list=[]\n",
    "    for dis in list(dis_dict.keys()):\n",
    "        mu = np.mean(proximal_dis_mortality[dis])\n",
    "        std = np.std(proximal_dis_mortality[dis])\n",
    "        z = (country_dis_mortality[country][dis]-mu)/std\n",
    "        country_proximal_zscore_list.append(z)\n",
    "        from scipy.stats import norm\n",
    "        ptemp = norm.sf(abs(z))\n",
    "    water_country_proximal_zscore_dict[country]=country_proximal_zscore_list\n",
    "\n",
    "\n",
    "water_country_distant_zscore_dict={}\n",
    "for country,dis_dict in new_water_country_distant_dis_dict.items():\n",
    "    country_distant_zscore_list=[]\n",
    "    for dis in list(dis_dict.keys()):\n",
    "        mu = np.mean(distant_dis_mortality[dis])\n",
    "        std = np.std(distant_dis_mortality[dis])\n",
    "        z = (country_dis_mortality[country][dis]-mu)/std\n",
    "        country_distant_zscore_list.append(z)\n",
    "        from scipy.stats import norm\n",
    "        ptemp = norm.sf(abs(z))\n",
    "    water_country_distant_zscore_dict[country]=country_distant_zscore_list\n",
    "\n",
    "water_pollution_zscore_df = pd.DataFrame()\n",
    "selected_countries=set(water_country_distant_zscore_dict.keys())\n",
    "tot_country_list=[]\n",
    "tot_zscore_list=[]\n",
    "location_network_list=[]\n",
    "for country in selected_countries:\n",
    "    farawaylist_cleaned = [x for x in water_country_distant_zscore_dict[country] if str(x) != 'nan']\n",
    "    proximallist_cleaned = [x for x in water_country_proximal_zscore_dict[country] if str(x) != 'nan']\n",
    "    tot_zscorelist_partial=farawaylist_cleaned+proximallist_cleaned\n",
    "    for i in proximallist_cleaned:\n",
    "        location_network_list.append(\"proximal\")\n",
    "        tot_zscore_list.append(i)\n",
    "        tot_country_list.append(country)\n",
    "    for i in farawaylist_cleaned:\n",
    "        location_network_list.append(\"distant\")\n",
    "        tot_zscore_list.append(i)\n",
    "        tot_country_list.append(country)\n",
    "\n",
    "water_pollution_zscore_df['location']=tot_country_list\n",
    "water_pollution_zscore_df['z-score']=tot_zscore_list\n",
    "water_pollution_zscore_df['network_location']=location_network_list\n",
    "\n",
    "for country in water_country_proximal_zscore_dict.keys():\n",
    "    try:\n",
    "        proximallist_cleaned = [x for x in water_country_proximal_zscore_dict[country] if str(x) != 'nan']\n",
    "        distantlist_cleaned = [x for x in water_country_distant_zscore_dict[country] if str(x) != 'nan']\n",
    "        prox_mean = np.mean(proximallist_cleaned)\n",
    "        distant_mean = np.mean(distantlist_cleaned)\n",
    "        pval_noprox = scipy.stats.mannwhitneyu(proximallist_cleaned, distantlist_cleaned, use_continuity=True, alternative='greater')[1]\n",
    "        print(country,prox_mean,distant_mean,round(pval_noprox,4),round(pval_distant,4))\n",
    "    except:\n",
    "        print(country,prox_mean,distant_mean,round(pval_noprox,4))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "4374d591-4d0b-40d3-98a2-7b022d7ba455",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "fig, ax = plt.subplots(figsize=(15, 20))\n",
    "sns.set(style=\"white\")\n",
    "g = sns.boxplot(x=\"z-score\", y=\"location\", orient=\"h\",\n",
    "                hue=\"network_location\", palette=[\"#2B7A98\", \"#54CAF2\"], order=my_order,\n",
    "                data=water_pollution_zscore_df)\n",
    "fontsize = 36\n",
    "plt.legend(loc=\"upper right\", frameon=True, fontsize=fontsize)\n",
    "plt.yticks(fontsize=36)\n",
    "\n",
    "for ax in plt.gcf().axes:\n",
    "    l = ax.get_xlabel()\n",
    "    ax.set_xlabel(l, fontsize=36)\n",
    "    ax.set_ylabel(l, fontsize=36)\n",
    "\n",
    "plt.ylabel('', fontsize=25)\n",
    "plt.xlabel('', fontsize=25)\n",
    "ax.spines['top'].set_visible(False)\n",
    "ax.spines['right'].set_visible(False)\n",
    "ax.spines['bottom'].set_visible(True)\n",
    "ax.spines['left'].set_visible(True)\n",
    "ax.set_xticklabels(ax.get_xticks(), size=36)\n",
    "\n",
    "pv_list = []  # Example p-values\n",
    "locations = []  # Example y-axis categories\n",
    "stars = []  # Based on the p-values, this is what we want to add\n",
    "\n",
    "\n",
    "\n",
    "for country in my_order:\n",
    "    proximallist_cleaned = [x for x in water_country_proximal_zscore_dict[country] if str(x) != 'nan']\n",
    "    distantlist_cleaned = [x for x in water_country_distant_zscore_dict[country] if str(x) != 'nan']\n",
    "    prox_mean = np.mean(proximallist_cleaned)\n",
    "    distant_mean = np.mean(distantlist_cleaned)\n",
    "    pval_noprox = scipy.stats.mannwhitneyu(proximallist_cleaned, distantlist_cleaned, use_continuity=True, alternative='greater')[1]\n",
    "    pv_list.append(pval_noprox)\n",
    "    locations.append(country)\n",
    "    if pval_noprox<0.001:\n",
    "        stars.append(\"***\")\n",
    "    elif pval_noprox>0.001 and pval_noprox<0.01:\n",
    "        stars.append(\"**\")\n",
    "    elif pval_noprox>0.01 and pval_noprox<0.05:\n",
    "        stars.append(\"*\")\n",
    "    else:\n",
    "        stars.append(\"n.s.\")\n",
    "\n",
    "\n",
    "for pv, loc, star in zip(pv_list, locations, stars):\n",
    "    yloc = my_order.index(loc) + 0.3  # This presumes 'my_order' list contains the order of y-categories\n",
    "    xloc = max(water_pollution_zscore_df[water_pollution_zscore_df['location'] == loc]['z-score']) + 0.5  \n",
    "    # Add the annotation\n",
    "    ax.text(xloc, yloc, star, fontsize=fontsize, color=\"black\", ha='center')\n",
    "\n",
    "plt.savefig('output/Figures/SD13D_v01_post_rev.pdf',format='pdf',dpi=300, bbox_inches = \"tight\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "0ab01d2e-5c1b-4e84-8721-5bdffd55b9c7",
   "metadata": {},
   "outputs": [],
   "source": [
    "#daly\n",
    "\n",
    "water_country_proximal_zscore_dict={}\n",
    "for country,dis_dict in country_water_chem_id_gbd_mean_dict.items():\n",
    "    country_proximal_zscore_list=[]\n",
    "    for dis in list(dis_dict.keys()):\n",
    "        mu = np.mean(proximal_dis_daly[dis])\n",
    "        std = np.std(proximal_dis_daly[dis])\n",
    "        z = (country_dis_DALYs[country][dis]-mu)/std\n",
    "        country_proximal_zscore_list.append(z)\n",
    "        from scipy.stats import norm\n",
    "        ptemp = norm.sf(abs(z))\n",
    "    water_country_proximal_zscore_dict[country]=country_proximal_zscore_list\n",
    "\n",
    "\n",
    "water_country_distant_zscore_dict={}\n",
    "for country,dis_dict in new_water_country_distant_dis_dict.items():\n",
    "    country_distant_zscore_list=[]\n",
    "    for dis in list(dis_dict.keys()):\n",
    "        mu = np.mean(distant_dis_daly[dis])\n",
    "        std = np.std(distant_dis_daly[dis])\n",
    "        z = (country_dis_DALYs[country][dis]-mu)/std\n",
    "        country_distant_zscore_list.append(z)\n",
    "        from scipy.stats import norm\n",
    "        ptemp = norm.sf(abs(z))\n",
    "    water_country_distant_zscore_dict[country]=country_distant_zscore_list\n",
    "\n",
    "water_pollution_zscore_df = pd.DataFrame()\n",
    "selected_countries=set(water_country_distant_zscore_dict.keys())\n",
    "tot_country_list=[]\n",
    "tot_zscore_list=[]\n",
    "location_network_list=[]\n",
    "for country in selected_countries:\n",
    "    farawaylist_cleaned = [x for x in water_country_distant_zscore_dict[country] if str(x) != 'nan']\n",
    "    proximallist_cleaned = [x for x in water_country_proximal_zscore_dict[country] if str(x) != 'nan']\n",
    "    tot_zscorelist_partial=farawaylist_cleaned+proximallist_cleaned\n",
    "    for i in proximallist_cleaned:\n",
    "        location_network_list.append(\"proximal\")\n",
    "        tot_zscore_list.append(i)\n",
    "        tot_country_list.append(country)\n",
    "    for i in farawaylist_cleaned:\n",
    "        location_network_list.append(\"distant\")\n",
    "        tot_zscore_list.append(i)\n",
    "        tot_country_list.append(country)\n",
    "\n",
    "water_pollution_zscore_df['location']=tot_country_list\n",
    "water_pollution_zscore_df['z-score']=tot_zscore_list\n",
    "water_pollution_zscore_df['network_location']=location_network_list\n",
    "\n",
    "for country in water_country_proximal_zscore_dict.keys():\n",
    "    try:\n",
    "        proximallist_cleaned = [x for x in water_country_proximal_zscore_dict[country] if str(x) != 'nan']\n",
    "        distantlist_cleaned = [x for x in water_country_distant_zscore_dict[country] if str(x) != 'nan']\n",
    "        prox_mean = np.mean(proximallist_cleaned)\n",
    "        distant_mean = np.mean(distantlist_cleaned)\n",
    "        pval_noprox = scipy.stats.mannwhitneyu(proximallist_cleaned, distantlist_cleaned, use_continuity=True, alternative='greater')[1]\n",
    "        print(country,prox_mean,distant_mean,round(pval_noprox,4),round(pval_distant,4))\n",
    "    except:\n",
    "        print(country,prox_mean,distant_mean,round(pval_noprox,4))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "73f16a00-9bb2-4fba-900e-75bccd22fe2d",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "fig, ax = plt.subplots(figsize=(15, 20))\n",
    "sns.set(style=\"white\")\n",
    "g = sns.boxplot(x=\"z-score\", y=\"location\", orient=\"h\",\n",
    "                hue=\"network_location\", palette=[\"#2B7A98\", \"#54CAF2\"], order=my_order,\n",
    "                data=water_pollution_zscore_df)\n",
    "fontsize = 36\n",
    "plt.legend(loc=\"upper right\", frameon=True, fontsize=fontsize)\n",
    "plt.yticks(fontsize=36)\n",
    "\n",
    "for ax in plt.gcf().axes:\n",
    "    l = ax.get_xlabel()\n",
    "    ax.set_xlabel(l, fontsize=36)\n",
    "    ax.set_ylabel(l, fontsize=36)\n",
    "\n",
    "plt.ylabel('', fontsize=25)\n",
    "plt.xlabel('', fontsize=25)\n",
    "ax.spines['top'].set_visible(False)\n",
    "ax.spines['right'].set_visible(False)\n",
    "ax.spines['bottom'].set_visible(True)\n",
    "ax.spines['left'].set_visible(True)\n",
    "ax.set_xticklabels(ax.get_xticks(), size=36)\n",
    "\n",
    "pv_list = []  # Example p-values\n",
    "locations = []  # Example y-axis categories\n",
    "stars = []  # Based on the p-values, this is what we want to add\n",
    "\n",
    "\n",
    "\n",
    "for country in my_order:\n",
    "    proximallist_cleaned = [x for x in water_country_proximal_zscore_dict[country] if str(x) != 'nan']\n",
    "    distantlist_cleaned = [x for x in water_country_distant_zscore_dict[country] if str(x) != 'nan']\n",
    "    prox_mean = np.mean(proximallist_cleaned)\n",
    "    distant_mean = np.mean(distantlist_cleaned)\n",
    "    pval_noprox = scipy.stats.mannwhitneyu(proximallist_cleaned, distantlist_cleaned, use_continuity=True, alternative='greater')[1]\n",
    "    pv_list.append(pval_noprox)\n",
    "    locations.append(country)\n",
    "    if pval_noprox<0.001:\n",
    "        stars.append(\"***\")\n",
    "    elif pval_noprox>0.001 and pval_noprox<0.01:\n",
    "        stars.append(\"**\")\n",
    "    elif pval_noprox>0.01 and pval_noprox<0.05:\n",
    "        stars.append(\"*\")\n",
    "    else:\n",
    "        stars.append(\"n.s.\")\n",
    "\n",
    "\n",
    "for pv, loc, star in zip(pv_list, locations, stars):\n",
    "    yloc = my_order.index(loc) + 0.3  # This presumes 'my_order' list contains the order of y-categories\n",
    "    xloc = max(water_pollution_zscore_df[water_pollution_zscore_df['location'] == loc]['z-score']) + 0.5  \n",
    "    # Add the annotation\n",
    "    ax.text(xloc, yloc, star, fontsize=fontsize, color=\"black\", ha='center')\n",
    "\n",
    "plt.savefig('output/Figures/SD14D_v01_post_rev.pdf',format='pdf',dpi=300, bbox_inches = \"tight\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "73c92ee6-c1b6-4ecd-aa1c-365c6670e56a",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "fd207cb3-0913-47f9-944a-974043c540dd",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "e6d20fe9-7915-4abd-8979-064a8a463c6d",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "374d8364-38d9-4cfa-b799-17e248e7ffdd",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c6852e8b-0767-4f0d-b433-5d1e3048e850",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "3718cfd3-fd1a-49dd-9784-2f3a3405247e",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "0b846444-3e90-46a6-b68d-dbc58e8bbd72",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "5bac6beb-ee3b-4ec6-9574-9b709415f05c",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "9f9eccec-ec54-4cb8-98aa-5a6b2aedfd83",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "55b3875a-a2ca-4835-ae88-85f8fdfbb980",
   "metadata": {},
   "outputs": [],
   "source": [
    "#LOW"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "74dd0811-5f99-4f83-960a-2dfabf39be66",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Proximal\n",
    "water_low_country_selected_dis_mean_prox_dict = {}\n",
    "for country,disset in water_low_country_selected_dis_dict.items():\n",
    "    dis_dict = {}\n",
    "    for dis in disset:\n",
    "        z_list = [] \n",
    "        if len(water_low_country_selected_chem_id_dict[country])>0:\n",
    "            for chem in water_low_country_selected_chem_id_dict[country]:\n",
    "                try:\n",
    "                    z_list.append(real_significant_final_exp_proximity[chem,dis][1])\n",
    "                except:\n",
    "                    pass\n",
    "        dis_dict[dis]= np.round(np.mean(z_list),2)            \n",
    "    water_low_country_selected_dis_mean_prox_dict[country] = dis_dict\n",
    "\n",
    "water_low_country_selected_chem_id_gbd_dict = {}\n",
    "for country,dis_dict in water_low_country_selected_dis_mean_prox_dict.items():\n",
    "    if len(dis_dict)>0:\n",
    "        dis_z_list = []\n",
    "        for dis,z in dis_dict.items():\n",
    "            try:\n",
    "                dis_z_list.append([water_disgenet_gbd_dict[dis],z])\n",
    "            except:\n",
    "                pass\n",
    "        water_low_country_selected_chem_id_gbd_dict[country] = dis_z_list\n",
    "\n",
    "water_low_country_selected_chem_id_gbd_mean_dict = {}\n",
    "for country,dislist in water_low_country_selected_chem_id_gbd_dict.items():\n",
    "    dis_dict = {}\n",
    "    dis_list = []\n",
    "    for el in dislist:\n",
    "        dis_list.append(el[0])\n",
    "    dis_list_counter_dict = dict(Counter(dis_list).most_common())\n",
    "    for dis,count in dis_list_counter_dict.items():\n",
    "        z_list = []\n",
    "        if count>1:\n",
    "            for el in dislist:\n",
    "                if dis==el[0]:\n",
    "                    z_list.append(el[1])\n",
    "            dis_dict[dis]=np.round(np.mean(z_list),2)\n",
    "        else:\n",
    "            for el in dislist:\n",
    "                if dis==el[0]:\n",
    "                    dis_dict[el[0]] = el[1]\n",
    "    water_low_country_selected_chem_id_gbd_mean_dict[country]=dis_dict\n",
    "\n",
    "proximal_dis_list=[] \n",
    "for country, disdict in water_low_country_selected_chem_id_gbd_mean_dict.items():\n",
    "    proximal_dis_list.extend(list(disdict.keys()))\n",
    "    \n",
    "proximal_dis_incidence={}\n",
    "country_set=set(GBD_df['location'])\n",
    "for dis in set(proximal_dis_list):\n",
    "    incidence_list=[]\n",
    "    for country in country_set:\n",
    "        incidence_list.append(country_dis_incidence[country][dis])\n",
    "    proximal_dis_incidence[dis]=incidence_list\n",
    "\n",
    "\n",
    "proximal_dis_prevalence={}\n",
    "for dis in set(proximal_dis_list):\n",
    "    prevalence_list=[]\n",
    "    for country in country_set:\n",
    "        prevalence_list.append(country_dis_prevalence[country][dis])\n",
    "    proximal_dis_prevalence[dis]=prevalence_list\n",
    "\n",
    "proximal_dis_mortality={}\n",
    "for dis in set(proximal_dis_list):\n",
    "    mortality_list=[]\n",
    "    for country in country_set:\n",
    "        mortality_list.append(country_dis_mortality[country][dis])\n",
    "    proximal_dis_mortality[dis]=mortality_list\n",
    "\n",
    "proximal_dis_daly={}\n",
    "for dis in set(proximal_dis_list):\n",
    "    daly_list=[]\n",
    "    for country in country_set:\n",
    "        daly_list.append(country_dis_DALYs[country][dis])\n",
    "    proximal_dis_daly[dis]=daly_list\n",
    "\n",
    "\n",
    "water_low_country_proximal_zscore_dict={}\n",
    "for country,dis_dict in water_low_country_selected_chem_id_gbd_mean_dict.items():\n",
    "    country_proximal_zscore_list=[]\n",
    "    for dis in list(dis_dict.keys()):\n",
    "        mu = np.mean(proximal_dis_incidence[dis])\n",
    "        std = np.std(proximal_dis_incidence[dis])\n",
    "        z = (country_dis_incidence[country][dis]-mu)/std\n",
    "        country_proximal_zscore_list.append(z)\n",
    "        from scipy.stats import norm\n",
    "        ptemp = norm.sf(abs(z))\n",
    "    if len(country_proximal_zscore_list)>0:\n",
    "        water_low_country_proximal_zscore_dict[country]=country_proximal_zscore_list\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "0eea5ea0-23b3-46ef-aafb-28fbbf941984",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Distant\n",
    "low_country_water_faraway_dis_dict={}\n",
    "for country,chemset in water_low_country_selected_chem_id_dict.items():\n",
    "    if len(chemset)>0:\n",
    "        far_away_dis_set = []\n",
    "        for chem in chemset:\n",
    "            for chem_dis_pair in faraway_exp_dis_dict.keys():\n",
    "                if chem in chem_dis_pair[0]:\n",
    "                    far_away_dis_set.append(chem_dis_pair[1])\n",
    "        low_country_water_faraway_dis_dict[country] = set(far_away_dis_set)\n",
    "\n",
    "low_country_water_faraway_dis_dict = {}\n",
    "for country,dis_dict in water_low_country_selected_chem_id_gbd_mean_dict.items():\n",
    "    chem_set = water_low_country_selected_chem_id_dict[country]\n",
    "    country_faraway_dis_list = []\n",
    "    country_specific_faraway_dis_dict = {}\n",
    "    for dis in faraway_dis_list:\n",
    "        if dis not in dis_dict.keys():\n",
    "            country_faraway_dis_list.append(dis)\n",
    "    for dis in country_faraway_dis_list:\n",
    "        if dis not in faraway_manually_added_diseases:\n",
    "            disgenet_name_list = []\n",
    "            z_score_list = []\n",
    "            for disgenet_name,gbd_id in n_long_faraway_water_disset_gbd_dict_cleaned_v2.items():\n",
    "                if dis == gbd_id:\n",
    "                    disgenet_name_list.append(disgenet_name)\n",
    "                    for chem in chem_set:\n",
    "                        try:\n",
    "                            z_score_list.append(exp_proximity_not_significant[chem,disgenet_name][1])\n",
    "                        except:\n",
    "                            pass\n",
    "            if np.mean(z_score_list)>1.5:\n",
    "                country_specific_faraway_dis_dict[dis]=np.mean(z_score_list)\n",
    "        elif dis=='Congenital heart anomalies':\n",
    "            disgenet_name_list = ['Congenital Heart Defects','Deafness, Congenital Heart Defects, and Posterior Embryotoxon',\n",
    "                                 'Congenital long QT syndrome']\n",
    "            z_score_list = []\n",
    "            for disgenet_name in disgenet_name_list:\n",
    "                for chem in chem_set:\n",
    "                    try:\n",
    "                        z_score_list.append(exp_proximity_not_significant[chem,disgenet_name][1])\n",
    "                    except:\n",
    "                        pass\n",
    "            if np.mean(z_score_list)>1.5:\n",
    "                country_specific_faraway_dis_dict[dis]=np.mean(z_score_list)\n",
    "        \n",
    "        elif dis=='Congenital musculoskeletal and limb anomalies':\n",
    "            disgenet_name_list = ['Congenital Fiber Type Disproportion','Myopathy, Actin, Congenital, with Excess of Thin Myofilaments',\n",
    "                                 'Myopathy, Actin, Congenital, With Cores','Muscular Dystrophy, Congenital, Megaconial Type',\n",
    "                                 'Myotonia Congenita','Myasthenic Syndromes, Congenital','Muscular Dystrophy, Congenital, Megaconial Type',\n",
    "                                 'Congenital Myasthenic Syndromes, Postsynaptic','Congenital Myasthenic Syndromes, Presynaptic',\n",
    "                                 'Muscular Dystrophy, Congenital, Due To Integrin Alpha-7 Deficiency','Muscular Dystrophy, Congenital, Type 1D',\n",
    "                                 'Congenital Fibrosis of the Extraocular Muscles']\n",
    "            z_score_list = []\n",
    "            for disgenet_name in disgenet_name_list:\n",
    "                for chem in chem_set:\n",
    "                    try:\n",
    "                        z_score_list.append(exp_proximity_not_significant[chem,disgenet_name][1])\n",
    "                    except:\n",
    "                        pass\n",
    "            if np.mean(z_score_list)>1.5:\n",
    "                country_specific_faraway_dis_dict[dis]=np.mean(z_score_list)\n",
    "            \n",
    "        elif dis=='Congenital birth defects':\n",
    "            disgenet_name_list = [dis for dis in list(diseases_genes_associated_ppi.keys()) if 'Congenital' in dis or 'congenital' in dis]\n",
    "            z_score_list = []\n",
    "            for disgenet_name in disgenet_name_list:\n",
    "                for chem in chem_set:\n",
    "                    try:\n",
    "                        z_score_list.append(exp_proximity_not_significant[chem,disgenet_name][1])\n",
    "                    except:\n",
    "                        pass\n",
    "            if np.mean(z_score_list)>1.5:\n",
    "                country_specific_faraway_dis_dict[dis]=np.mean(z_score_list)\n",
    "\n",
    "        elif dis=='Turner syndrome':\n",
    "            disgenet_name_list = [dis for dis in list(diseases_genes_associated_ppi.keys()) if 'Turner' in dis]\n",
    "            z_score_list = []\n",
    "            for disgenet_name in disgenet_name_list:\n",
    "                for chem in chem_set:\n",
    "                    try:\n",
    "                        z_score_list.append(exp_proximity_not_significant[chem,disgenet_name][1])\n",
    "                    except:\n",
    "                        pass\n",
    "            if np.mean(z_score_list)>1.5:\n",
    "                country_specific_faraway_dis_dict[dis]=np.mean(z_score_list)\n",
    "    low_country_water_faraway_dis_dict[country] = country_specific_faraway_dis_dict\n",
    "\n",
    "\n",
    "low_country_noprox_water_dis_dict = {}\n",
    "for country,dis_dict in water_low_country_selected_chem_id_gbd_mean_dict.items():\n",
    "    n_dis = len(dis_dict)\n",
    "    chem_set = water_low_country_selected_chem_id_dict[country]\n",
    "    country_noprox_dis_list = []\n",
    "    country_specific_noprox_dis_dict = {}\n",
    "    \n",
    "    for gbd_dis,dislist in no_prox_air_disset_gbd_dict_cleaned_inv.items():\n",
    "        if gbd_dis not in dis_dict.keys():\n",
    "            if gbd_dis not in list(low_country_water_faraway_dis_dict[country].keys()):\n",
    "                country_noprox_dis_list.append(gbd_dis)\n",
    "    for dis in country_noprox_dis_list:\n",
    "        disgenet_name_list = no_prox_air_disset_gbd_dict_cleaned_inv[dis]\n",
    "        z_score_list = []\n",
    "        for chem in chem_set:\n",
    "            for disgenet_name in disgenet_name_list:\n",
    "                try:\n",
    "                    z_score_list.append(exp_proximity_not_significant[chem,disgenet_name][1])\n",
    "                except:\n",
    "                    pass\n",
    "        if np.mean(z_score_list)>-1.5:\n",
    "            country_specific_noprox_dis_dict[dis]=np.mean(z_score_list)\n",
    "    country_specific_noprox_dis_dict_sort = {k: v for k, v in sorted(country_specific_noprox_dis_dict.items(), key=lambda item: item[1])}\n",
    "    \n",
    "    if len(country_specific_noprox_dis_dict_sort)>n_dis:\n",
    "        country_specific_noprox_dis_dict_sort_keys = list(country_specific_noprox_dis_dict_sort.keys())\n",
    "        n_country_specific_noprox_dis_dict_sort = {}\n",
    "        for k in list(country_specific_noprox_dis_dict_sort_keys)[:n_dis]:\n",
    "            n_country_specific_noprox_dis_dict_sort[k]=country_specific_noprox_dis_dict_sort[k]\n",
    "        \n",
    "        low_country_noprox_water_dis_dict[country] = n_country_specific_noprox_dis_dict_sort\n",
    "    else:\n",
    "        low_country_noprox_water_dis_dict[country] = country_specific_noprox_dis_dict_sort\n",
    "\n",
    "#Let's just merge the no-proximal dictionary and the faraway dictionary\n",
    "new_low_water_country_distant_dis_dict = {}\n",
    "for country,dis_dict in water_low_country_selected_chem_id_gbd_mean_dict.items():\n",
    "    n_proxdis = len(dis_dict)\n",
    "    n_distantdis = len(low_country_water_faraway_dis_dict[country])\n",
    "    noprox_list = list(low_country_noprox_water_dis_dict[country].keys())\n",
    "    missing_dis = n_proxdis-n_distantdis\n",
    "    try:\n",
    "        dislist = noprox_list[:missing_dis]\n",
    "        distant_dict = low_country_water_faraway_dis_dict[country].copy()\n",
    "        for dis in dislist:\n",
    "            distant_dict[dis] = low_country_noprox_water_dis_dict[country][dis]\n",
    "        new_low_water_country_distant_dis_dict[country] = distant_dict\n",
    "    except:\n",
    "        distant_dict = {**low_country_water_faraway_dis_dict[country], **low_country_noprox_water_dis_dict[country]}\n",
    "        new_low_water_country_distant_dis_dict[country] = distant_dict\n",
    "\n",
    "distant_dis_list=[] \n",
    "for country, disdict in new_low_water_country_distant_dis_dict.items():\n",
    "    distant_dis_list.extend(list(disdict.keys()))\n",
    "    \n",
    "distant_dis_incidence={}\n",
    "country_set=set(GBD_df['location'])\n",
    "for dis in set(distant_dis_list):\n",
    "    incidence_list=[]\n",
    "    for country in country_set:\n",
    "        incidence_list.append(country_dis_incidence[country][dis])\n",
    "    distant_dis_incidence[dis]=incidence_list\n",
    "\n",
    "\n",
    "distant_dis_prevalence={}\n",
    "for dis in set(distant_dis_list):\n",
    "     prevalence_list=[]\n",
    "     for country in country_set:\n",
    "         prevalence_list.append(country_dis_prevalence[country][dis])\n",
    "     distant_dis_prevalence[dis]=prevalence_list\n",
    "\n",
    "distant_dis_mortality={}\n",
    "for dis in set(distant_dis_list):\n",
    "     mortality_list=[]\n",
    "     for country in country_set:\n",
    "         mortality_list.append(country_dis_mortality[country][dis])\n",
    "     distant_dis_mortality[dis]=mortality_list\n",
    "\n",
    "distant_dis_daly={}\n",
    "for dis in set(distant_dis_list):\n",
    "     daly_list=[]\n",
    "     for country in country_set:\n",
    "         daly_list.append(country_dis_DALYs[country][dis])\n",
    "     distant_dis_daly[dis]=daly_list\n",
    "\n",
    "water_low_country_distant_zscore_dict={}\n",
    "for country,dis_dict in new_low_water_country_distant_dis_dict.items():\n",
    "    country_distant_zscore_list=[]\n",
    "    for dis in list(dis_dict.keys()):\n",
    "        mu = np.mean(distant_dis_incidence[dis])\n",
    "        std = np.std(distant_dis_incidence[dis])\n",
    "        z = (country_dis_incidence[country][dis]-mu)/std\n",
    "        country_distant_zscore_list.append(z)\n",
    "\n",
    "        from scipy.stats import norm\n",
    "        ptemp = norm.sf(abs(z))\n",
    "    water_low_country_distant_zscore_dict[country]=country_distant_zscore_list"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "5b58cd9c-91e9-413b-b3eb-cb237077e106",
   "metadata": {},
   "outputs": [],
   "source": [
    "water_pollution_low_zscore_df = pd.DataFrame()\n",
    "\n",
    "selected_countries=set(water_low_country_proximal_zscore_dict.keys())\n",
    "tot_country_list=[]\n",
    "tot_zscore_list=[]\n",
    "location_network_list=[]\n",
    "for country in selected_countries:\n",
    "    farawaylist_cleaned = [x for x in water_low_country_distant_zscore_dict[country] if str(x) != 'nan']\n",
    "    proximallist_cleaned = [x for x in water_low_country_proximal_zscore_dict[country] if str(x) != 'nan']\n",
    "    tot_zscorelist_partial=farawaylist_cleaned+proximallist_cleaned\n",
    "    for i in proximallist_cleaned:\n",
    "        location_network_list.append(\"proximal\")\n",
    "        tot_zscore_list.append(i)\n",
    "        tot_country_list.append(country)\n",
    "    for i in farawaylist_cleaned:\n",
    "        location_network_list.append(\"distant\")\n",
    "        tot_zscore_list.append(i)\n",
    "        tot_country_list.append(country)\n",
    "\n",
    "water_pollution_low_zscore_df['location']=tot_country_list\n",
    "water_pollution_low_zscore_df['z-score']=tot_zscore_list\n",
    "water_pollution_low_zscore_df['network_location']=location_network_list\n",
    "\n",
    "my_order = sorted(list(set(water_pollution_low_zscore_df['location'])))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "2cef8153-6534-4c21-ab7c-220d233bbe59",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Let's test the difference between proximal and no-proximal (because with distant is kind of obvious and it doesn't\n",
    "#make so much sense)\n",
    "import scipy\n",
    "for country in water_low_country_proximal_zscore_dict.keys():\n",
    "    try:\n",
    "        proximallist_cleaned = [x for x in water_low_country_proximal_zscore_dict[country] if str(x) != 'nan']\n",
    "        distantlist_cleaned = [x for x in water_low_country_distant_zscore_dict[country] if str(x) != 'nan']\n",
    "        prox_mean = np.mean(proximallist_cleaned)\n",
    "        distant_mean = np.mean(distantlist_cleaned)\n",
    "        pval_noprox = scipy.stats.mannwhitneyu(proximallist_cleaned, distantlist_cleaned, use_continuity=True, alternative='greater')[1]\n",
    "        print(country,prox_mean,distant_mean,round(pval_noprox,4),round(pval_distant,4))\n",
    "    except:\n",
    "        print(country,prox_mean,distant_mean,round(pval_noprox,4))\n",
    "        "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d9ae8132-0db2-4dbe-9725-a29d07489cb9",
   "metadata": {},
   "outputs": [],
   "source": [
    "import scipy \n",
    "# Assuming your data loading and plot setup code here\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(15, 20))\n",
    "sns.set(style=\"white\")\n",
    "g = sns.boxplot(x=\"z-score\", y=\"location\", orient=\"h\",\n",
    "                hue=\"network_location\", palette=[\"#2B7A98\", \"#54CAF2\"], order=my_order,\n",
    "                data=water_pollution_low_zscore_df)\n",
    "fontsize = 36\n",
    "plt.legend(loc=\"upper right\", frameon=True, fontsize=fontsize)\n",
    "plt.yticks(fontsize=36)\n",
    "\n",
    "for ax in plt.gcf().axes:\n",
    "    l = ax.get_xlabel()\n",
    "    ax.set_xlabel(l, fontsize=36)\n",
    "    ax.set_ylabel(l, fontsize=36)\n",
    "\n",
    "plt.ylabel('', fontsize=25)\n",
    "plt.xlabel('', fontsize=25)\n",
    "ax.spines['top'].set_visible(False)\n",
    "ax.spines['right'].set_visible(False)\n",
    "ax.spines['bottom'].set_visible(True)\n",
    "ax.spines['left'].set_visible(True)\n",
    "ax.set_xticklabels(ax.get_xticks(), size=36)\n",
    "\n",
    "pv_list = [] \n",
    "locations = [] \n",
    "stars = [] \n",
    "\n",
    "\n",
    "for country in my_order:\n",
    "    proximallist_cleaned = [x for x in water_low_country_proximal_zscore_dict[country] if str(x) != 'nan']\n",
    "    distantlist_cleaned = [x for x in water_low_country_distant_zscore_dict[country] if str(x) != 'nan']\n",
    "    prox_mean = np.mean(proximallist_cleaned)\n",
    "    distant_mean = np.mean(distantlist_cleaned)\n",
    "    pval_noprox = scipy.stats.mannwhitneyu(proximallist_cleaned, distantlist_cleaned, use_continuity=True, alternative='greater')[1]\n",
    "    pv_list.append(pval_noprox)\n",
    "    locations.append(country)\n",
    "    if pval_noprox<0.001:\n",
    "        stars.append(\"***\")\n",
    "    elif pval_noprox>0.001 and pval_noprox<0.01:\n",
    "        stars.append(\"**\")\n",
    "    elif pval_noprox>0.01 and pval_noprox<0.05:\n",
    "        stars.append(\"*\")\n",
    "    else:\n",
    "        stars.append(\"n.s.\")\n",
    "\n",
    "for pv, loc, star in zip(pv_list, locations, stars):\n",
    "    yloc = my_order.index(loc) + 0.4  \n",
    "    xloc = max(water_pollution_low_zscore_df[water_pollution_low_zscore_df['location'] == loc]['z-score']) + 0.5  \n",
    "  \n",
    "    # Add the annotation\n",
    "    ax.text(xloc, yloc, star, fontsize=fontsize, color=\"black\", ha='center')\n",
    "\n",
    "plt.savefig('output/Figures/SD11C_v01_post_rev.pdf',format='pdf',dpi=300, bbox_inches = \"tight\")\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "0d9648df-fe0e-4af1-ad30-e4d83d6c4b27",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "50e1cb4f-b2a4-4829-8cdb-afdc67d44bd3",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "310fd3e0-bcb3-44c0-9252-12bd22236024",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Prevalence\n",
    "\n",
    "water_country_proximal_zscore_dict={}\n",
    "for country,dis_dict in water_low_country_selected_chem_id_gbd_mean_dict.items():\n",
    "    country_proximal_zscore_list=[]\n",
    "    for dis in list(dis_dict.keys()):\n",
    "        mu = np.mean(proximal_dis_prevalence[dis])\n",
    "        std = np.std(proximal_dis_prevalence[dis])\n",
    "        z = (country_dis_prevalence[country][dis]-mu)/std\n",
    "        country_proximal_zscore_list.append(z)\n",
    "        from scipy.stats import norm\n",
    "        ptemp = norm.sf(abs(z))\n",
    "    water_country_proximal_zscore_dict[country]=country_proximal_zscore_list\n",
    "\n",
    "\n",
    "water_country_distant_zscore_dict={}\n",
    "for country,dis_dict in new_low_water_country_distant_dis_dict.items():\n",
    "    country_distant_zscore_list=[]\n",
    "    for dis in list(dis_dict.keys()):\n",
    "        mu = np.mean(distant_dis_prevalence[dis])\n",
    "        std = np.std(distant_dis_prevalence[dis])\n",
    "        z = (country_dis_prevalence[country][dis]-mu)/std\n",
    "        country_distant_zscore_list.append(z)\n",
    "        from scipy.stats import norm\n",
    "        ptemp = norm.sf(abs(z))\n",
    "    water_country_distant_zscore_dict[country]=country_distant_zscore_list\n",
    "\n",
    "water_pollution_zscore_df = pd.DataFrame()\n",
    "selected_countries=set(water_country_distant_zscore_dict.keys())\n",
    "tot_country_list=[]\n",
    "tot_zscore_list=[]\n",
    "location_network_list=[]\n",
    "for country in selected_countries:\n",
    "    farawaylist_cleaned = [x for x in water_country_distant_zscore_dict[country] if str(x) != 'nan']\n",
    "    proximallist_cleaned = [x for x in water_country_proximal_zscore_dict[country] if str(x) != 'nan']\n",
    "    tot_zscorelist_partial=farawaylist_cleaned+proximallist_cleaned\n",
    "    for i in proximallist_cleaned:\n",
    "        location_network_list.append(\"proximal\")\n",
    "        tot_zscore_list.append(i)\n",
    "        tot_country_list.append(country)\n",
    "    for i in farawaylist_cleaned:\n",
    "        location_network_list.append(\"distant\")\n",
    "        tot_zscore_list.append(i)\n",
    "        tot_country_list.append(country)\n",
    "\n",
    "water_pollution_zscore_df['location']=tot_country_list\n",
    "water_pollution_zscore_df['z-score']=tot_zscore_list\n",
    "water_pollution_zscore_df['network_location']=location_network_list\n",
    "\n",
    "for country in water_country_proximal_zscore_dict.keys():\n",
    "    try:\n",
    "        proximallist_cleaned = [x for x in water_country_proximal_zscore_dict[country] if str(x) != 'nan']\n",
    "        distantlist_cleaned = [x for x in water_country_distant_zscore_dict[country] if str(x) != 'nan']\n",
    "        prox_mean = np.mean(proximallist_cleaned)\n",
    "        distant_mean = np.mean(distantlist_cleaned)\n",
    "        pval_noprox = scipy.stats.mannwhitneyu(proximallist_cleaned, distantlist_cleaned, use_continuity=True, alternative='greater')[1]\n",
    "        print(country,prox_mean,distant_mean,round(pval_noprox,4),round(pval_distant,4))\n",
    "    except:\n",
    "        print(country,prox_mean,distant_mean,round(pval_noprox,4))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "4aae58e4-0757-4799-98c7-d2e7f070cc10",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "fig, ax = plt.subplots(figsize=(15, 20))\n",
    "sns.set(style=\"white\")\n",
    "g = sns.boxplot(x=\"z-score\", y=\"location\", orient=\"h\",\n",
    "                hue=\"network_location\", palette=[\"#2B7A98\", \"#54CAF2\"], order=my_order,\n",
    "                data=water_pollution_zscore_df)\n",
    "fontsize = 36\n",
    "plt.legend(loc=\"upper right\", frameon=True, fontsize=fontsize)\n",
    "plt.yticks(fontsize=36)\n",
    "\n",
    "for ax in plt.gcf().axes:\n",
    "    l = ax.get_xlabel()\n",
    "    ax.set_xlabel(l, fontsize=36)\n",
    "    ax.set_ylabel(l, fontsize=36)\n",
    "\n",
    "plt.ylabel('', fontsize=25)\n",
    "plt.xlabel('', fontsize=25)\n",
    "ax.spines['top'].set_visible(False)\n",
    "ax.spines['right'].set_visible(False)\n",
    "ax.spines['bottom'].set_visible(True)\n",
    "ax.spines['left'].set_visible(True)\n",
    "ax.set_xticklabels(ax.get_xticks(), size=36)\n",
    "\n",
    "pv_list = []  # Example p-values\n",
    "locations = []  # Example y-axis categories\n",
    "stars = []  # Based on the p-values, this is what we want to add\n",
    "\n",
    "\n",
    "\n",
    "for country in my_order:\n",
    "    proximallist_cleaned = [x for x in water_country_proximal_zscore_dict[country] if str(x) != 'nan']\n",
    "    distantlist_cleaned = [x for x in water_country_distant_zscore_dict[country] if str(x) != 'nan']\n",
    "    prox_mean = np.mean(proximallist_cleaned)\n",
    "    distant_mean = np.mean(distantlist_cleaned)\n",
    "    pval_noprox = scipy.stats.mannwhitneyu(proximallist_cleaned, distantlist_cleaned, use_continuity=True, alternative='greater')[1]\n",
    "    pv_list.append(pval_noprox)\n",
    "    locations.append(country)\n",
    "    if pval_noprox<0.001:\n",
    "        stars.append(\"***\")\n",
    "    elif pval_noprox>0.001 and pval_noprox<0.01:\n",
    "        stars.append(\"**\")\n",
    "    elif pval_noprox>0.01 and pval_noprox<0.05:\n",
    "        stars.append(\"*\")\n",
    "    else:\n",
    "        stars.append(\"n.s.\")\n",
    "\n",
    "\n",
    "for pv, loc, star in zip(pv_list, locations, stars):\n",
    "    yloc = my_order.index(loc) + 0.3  # This presumes 'my_order' list contains the order of y-categories\n",
    "    xloc = max(water_pollution_zscore_df[water_pollution_zscore_df['location'] == loc]['z-score']) + 0.5  \n",
    "    # Add the annotation\n",
    "    ax.text(xloc, yloc, star, fontsize=fontsize, color=\"black\", ha='center')\n",
    "\n",
    "plt.savefig('output/Figures/SD12E_v01_post_rev.pdf',format='pdf',dpi=300, bbox_inches = \"tight\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "05473dd0-2d40-40ed-b76f-a34b75f167d6",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "0e97d067-3bee-41c8-b325-1ea718177bda",
   "metadata": {},
   "outputs": [],
   "source": [
    "#mortality\n",
    "\n",
    "water_country_proximal_zscore_dict={}\n",
    "for country,dis_dict in water_low_country_selected_chem_id_gbd_mean_dict.items():\n",
    "    country_proximal_zscore_list=[]\n",
    "    for dis in list(dis_dict.keys()):\n",
    "        mu = np.mean(proximal_dis_mortality[dis])\n",
    "        std = np.std(proximal_dis_mortality[dis])\n",
    "        z = (country_dis_mortality[country][dis]-mu)/std\n",
    "        country_proximal_zscore_list.append(z)\n",
    "        from scipy.stats import norm\n",
    "        ptemp = norm.sf(abs(z))\n",
    "    water_country_proximal_zscore_dict[country]=country_proximal_zscore_list\n",
    "\n",
    "\n",
    "water_country_distant_zscore_dict={}\n",
    "for country,dis_dict in new_low_water_country_distant_dis_dict.items():\n",
    "    country_distant_zscore_list=[]\n",
    "    for dis in list(dis_dict.keys()):\n",
    "        mu = np.mean(distant_dis_mortality[dis])\n",
    "        std = np.std(distant_dis_mortality[dis])\n",
    "        z = (country_dis_mortality[country][dis]-mu)/std\n",
    "        country_distant_zscore_list.append(z)\n",
    "        from scipy.stats import norm\n",
    "        ptemp = norm.sf(abs(z))\n",
    "    water_country_distant_zscore_dict[country]=country_distant_zscore_list\n",
    "\n",
    "water_pollution_zscore_df = pd.DataFrame()\n",
    "selected_countries=set(water_country_distant_zscore_dict.keys())\n",
    "tot_country_list=[]\n",
    "tot_zscore_list=[]\n",
    "location_network_list=[]\n",
    "for country in selected_countries:\n",
    "    farawaylist_cleaned = [x for x in water_country_distant_zscore_dict[country] if str(x) != 'nan']\n",
    "    proximallist_cleaned = [x for x in water_country_proximal_zscore_dict[country] if str(x) != 'nan']\n",
    "    tot_zscorelist_partial=farawaylist_cleaned+proximallist_cleaned\n",
    "    for i in proximallist_cleaned:\n",
    "        location_network_list.append(\"proximal\")\n",
    "        tot_zscore_list.append(i)\n",
    "        tot_country_list.append(country)\n",
    "    for i in farawaylist_cleaned:\n",
    "        location_network_list.append(\"distant\")\n",
    "        tot_zscore_list.append(i)\n",
    "        tot_country_list.append(country)\n",
    "\n",
    "water_pollution_zscore_df['location']=tot_country_list\n",
    "water_pollution_zscore_df['z-score']=tot_zscore_list\n",
    "water_pollution_zscore_df['network_location']=location_network_list\n",
    "\n",
    "for country in water_country_proximal_zscore_dict.keys():\n",
    "    try:\n",
    "        proximallist_cleaned = [x for x in water_country_proximal_zscore_dict[country] if str(x) != 'nan']\n",
    "        distantlist_cleaned = [x for x in water_country_distant_zscore_dict[country] if str(x) != 'nan']\n",
    "        prox_mean = np.mean(proximallist_cleaned)\n",
    "        distant_mean = np.mean(distantlist_cleaned)\n",
    "        pval_noprox = scipy.stats.mannwhitneyu(proximallist_cleaned, distantlist_cleaned, use_continuity=True, alternative='greater')[1]\n",
    "        print(country,prox_mean,distant_mean,round(pval_noprox,4),round(pval_distant,4))\n",
    "    except:\n",
    "        print(country,prox_mean,distant_mean,round(pval_noprox,4))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "2bb081a0-c6d6-4eef-9c02-f3d63251a70a",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "fig, ax = plt.subplots(figsize=(15, 20))\n",
    "sns.set(style=\"white\")\n",
    "g = sns.boxplot(x=\"z-score\", y=\"location\", orient=\"h\",\n",
    "                hue=\"network_location\", palette=[\"#2B7A98\", \"#54CAF2\"], order=my_order,\n",
    "                data=water_pollution_zscore_df)\n",
    "fontsize = 36\n",
    "plt.legend(loc=\"upper right\", frameon=True, fontsize=fontsize)\n",
    "plt.yticks(fontsize=36)\n",
    "\n",
    "for ax in plt.gcf().axes:\n",
    "    l = ax.get_xlabel()\n",
    "    ax.set_xlabel(l, fontsize=36)\n",
    "    ax.set_ylabel(l, fontsize=36)\n",
    "\n",
    "plt.ylabel('', fontsize=25)\n",
    "plt.xlabel('', fontsize=25)\n",
    "ax.spines['top'].set_visible(False)\n",
    "ax.spines['right'].set_visible(False)\n",
    "ax.spines['bottom'].set_visible(True)\n",
    "ax.spines['left'].set_visible(True)\n",
    "ax.set_xticklabels(ax.get_xticks(), size=36)\n",
    "\n",
    "pv_list = []  # Example p-values\n",
    "locations = []  # Example y-axis categories\n",
    "stars = []  # Based on the p-values, this is what we want to add\n",
    "\n",
    "\n",
    "\n",
    "for country in my_order:\n",
    "    proximallist_cleaned = [x for x in water_country_proximal_zscore_dict[country] if str(x) != 'nan']\n",
    "    distantlist_cleaned = [x for x in water_country_distant_zscore_dict[country] if str(x) != 'nan']\n",
    "    prox_mean = np.mean(proximallist_cleaned)\n",
    "    distant_mean = np.mean(distantlist_cleaned)\n",
    "    pval_noprox = scipy.stats.mannwhitneyu(proximallist_cleaned, distantlist_cleaned, use_continuity=True, alternative='greater')[1]\n",
    "    pv_list.append(pval_noprox)\n",
    "    locations.append(country)\n",
    "    if pval_noprox<0.001:\n",
    "        stars.append(\"***\")\n",
    "    elif pval_noprox>0.001 and pval_noprox<0.01:\n",
    "        stars.append(\"**\")\n",
    "    elif pval_noprox>0.01 and pval_noprox<0.05:\n",
    "        stars.append(\"*\")\n",
    "    else:\n",
    "        stars.append(\"n.s.\")\n",
    "\n",
    "\n",
    "for pv, loc, star in zip(pv_list, locations, stars):\n",
    "    yloc = my_order.index(loc) + 0.3  # This presumes 'my_order' list contains the order of y-categories\n",
    "    xloc = max(water_pollution_zscore_df[water_pollution_zscore_df['location'] == loc]['z-score']) + 0.5  \n",
    "    # Add the annotation\n",
    "    ax.text(xloc, yloc, star, fontsize=fontsize, color=\"black\", ha='center')\n",
    "\n",
    "plt.savefig('output/Figures/SD13E_v01_post_rev.pdf',format='pdf',dpi=300, bbox_inches = \"tight\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d4136c3f-447d-419b-93fc-fd95b901c956",
   "metadata": {},
   "outputs": [],
   "source": [
    "#daly\n",
    "\n",
    "water_country_proximal_zscore_dict={}\n",
    "for country,dis_dict in water_low_country_selected_chem_id_gbd_mean_dict.items():\n",
    "    country_proximal_zscore_list=[]\n",
    "    for dis in list(dis_dict.keys()):\n",
    "        mu = np.mean(proximal_dis_daly[dis])\n",
    "        std = np.std(proximal_dis_daly[dis])\n",
    "        z = (country_dis_DALYs[country][dis]-mu)/std\n",
    "        country_proximal_zscore_list.append(z)\n",
    "        from scipy.stats import norm\n",
    "        ptemp = norm.sf(abs(z))\n",
    "    water_country_proximal_zscore_dict[country]=country_proximal_zscore_list\n",
    "\n",
    "\n",
    "water_country_distant_zscore_dict={}\n",
    "for country,dis_dict in new_low_water_country_distant_dis_dict.items():\n",
    "    country_distant_zscore_list=[]\n",
    "    for dis in list(dis_dict.keys()):\n",
    "        mu = np.mean(distant_dis_daly[dis])\n",
    "        std = np.std(distant_dis_daly[dis])\n",
    "        z = (country_dis_DALYs[country][dis]-mu)/std\n",
    "        country_distant_zscore_list.append(z)\n",
    "        from scipy.stats import norm\n",
    "        ptemp = norm.sf(abs(z))\n",
    "    water_country_distant_zscore_dict[country]=country_distant_zscore_list\n",
    "\n",
    "water_pollution_zscore_df = pd.DataFrame()\n",
    "selected_countries=set(water_country_distant_zscore_dict.keys())\n",
    "tot_country_list=[]\n",
    "tot_zscore_list=[]\n",
    "location_network_list=[]\n",
    "for country in selected_countries:\n",
    "    farawaylist_cleaned = [x for x in water_country_distant_zscore_dict[country] if str(x) != 'nan']\n",
    "    proximallist_cleaned = [x for x in water_country_proximal_zscore_dict[country] if str(x) != 'nan']\n",
    "    tot_zscorelist_partial=farawaylist_cleaned+proximallist_cleaned\n",
    "    for i in proximallist_cleaned:\n",
    "        location_network_list.append(\"proximal\")\n",
    "        tot_zscore_list.append(i)\n",
    "        tot_country_list.append(country)\n",
    "    for i in farawaylist_cleaned:\n",
    "        location_network_list.append(\"distant\")\n",
    "        tot_zscore_list.append(i)\n",
    "        tot_country_list.append(country)\n",
    "\n",
    "water_pollution_zscore_df['location']=tot_country_list\n",
    "water_pollution_zscore_df['z-score']=tot_zscore_list\n",
    "water_pollution_zscore_df['network_location']=location_network_list\n",
    "\n",
    "for country in water_country_proximal_zscore_dict.keys():\n",
    "    try:\n",
    "        proximallist_cleaned = [x for x in water_country_proximal_zscore_dict[country] if str(x) != 'nan']\n",
    "        distantlist_cleaned = [x for x in water_country_distant_zscore_dict[country] if str(x) != 'nan']\n",
    "        prox_mean = np.mean(proximallist_cleaned)\n",
    "        distant_mean = np.mean(distantlist_cleaned)\n",
    "        pval_noprox = scipy.stats.mannwhitneyu(proximallist_cleaned, distantlist_cleaned, use_continuity=True, alternative='greater')[1]\n",
    "        print(country,prox_mean,distant_mean,round(pval_noprox,4),round(pval_distant,4))\n",
    "    except:\n",
    "        print(country,prox_mean,distant_mean,round(pval_noprox,4))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "7ec5f8dc-d4b1-4ac6-842e-935314414682",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "fig, ax = plt.subplots(figsize=(15, 20))\n",
    "sns.set(style=\"white\")\n",
    "g = sns.boxplot(x=\"z-score\", y=\"location\", orient=\"h\",\n",
    "                hue=\"network_location\", palette=[\"#2B7A98\", \"#54CAF2\"], order=my_order,\n",
    "                data=water_pollution_zscore_df)\n",
    "fontsize = 36\n",
    "plt.legend(loc=\"upper right\", frameon=True, fontsize=fontsize)\n",
    "plt.yticks(fontsize=36)\n",
    "\n",
    "for ax in plt.gcf().axes:\n",
    "    l = ax.get_xlabel()\n",
    "    ax.set_xlabel(l, fontsize=36)\n",
    "    ax.set_ylabel(l, fontsize=36)\n",
    "\n",
    "plt.ylabel('', fontsize=25)\n",
    "plt.xlabel('', fontsize=25)\n",
    "ax.spines['top'].set_visible(False)\n",
    "ax.spines['right'].set_visible(False)\n",
    "ax.spines['bottom'].set_visible(True)\n",
    "ax.spines['left'].set_visible(True)\n",
    "ax.set_xticklabels(ax.get_xticks(), size=36)\n",
    "\n",
    "pv_list = []  # Example p-values\n",
    "locations = []  # Example y-axis categories\n",
    "stars = []  # Based on the p-values, this is what we want to add\n",
    "\n",
    "\n",
    "\n",
    "for country in my_order:\n",
    "    proximallist_cleaned = [x for x in water_country_proximal_zscore_dict[country] if str(x) != 'nan']\n",
    "    distantlist_cleaned = [x for x in water_country_distant_zscore_dict[country] if str(x) != 'nan']\n",
    "    prox_mean = np.mean(proximallist_cleaned)\n",
    "    distant_mean = np.mean(distantlist_cleaned)\n",
    "    pval_noprox = scipy.stats.mannwhitneyu(proximallist_cleaned, distantlist_cleaned, use_continuity=True, alternative='greater')[1]\n",
    "    pv_list.append(pval_noprox)\n",
    "    locations.append(country)\n",
    "    if pval_noprox<0.001:\n",
    "        stars.append(\"***\")\n",
    "    elif pval_noprox>0.001 and pval_noprox<0.01:\n",
    "        stars.append(\"**\")\n",
    "    elif pval_noprox>0.01 and pval_noprox<0.05:\n",
    "        stars.append(\"*\")\n",
    "    else:\n",
    "        stars.append(\"n.s.\")\n",
    "\n",
    "\n",
    "for pv, loc, star in zip(pv_list, locations, stars):\n",
    "    yloc = my_order.index(loc) + 0.3  # This presumes 'my_order' list contains the order of y-categories\n",
    "    xloc = max(water_pollution_zscore_df[water_pollution_zscore_df['location'] == loc]['z-score']) + 0.5  \n",
    "    # Add the annotation\n",
    "    ax.text(xloc, yloc, star, fontsize=fontsize, color=\"black\", ha='center')\n",
    "\n",
    "plt.savefig('output/Figures/SD14E_v01_post_rev.pdf',format='pdf',dpi=300, bbox_inches = \"tight\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d4178559-1048-4a1a-b2fc-45fe20cb3f0e",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "03dcfbc1-9b68-492d-ac22-e86789a018aa",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "cf6d8ad8-bdb8-422e-bbab-bb1b56029e91",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "411901ef-d04a-45a6-878f-f311f4b6d691",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "44fcb43f-892f-44b0-a6be-29ea837f062a",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "0b5a75f5-8f6d-44e1-a38b-4f7d4bf91aa6",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "cdd6ee4e-2493-46b1-885d-e24ec88a7335",
   "metadata": {},
   "outputs": [],
   "source": [
    "#HIGH\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "3cb1476a-312a-4f13-be38-88c27a09fff1",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Proximal\n",
    "water_high_country_selected_dis_mean_prox_dict = {}\n",
    "for country,disset in water_high_country_selected_dis_dict.items():\n",
    "    dis_dict = {}\n",
    "    for dis in disset:\n",
    "        z_list = [] \n",
    "        if len(water_high_country_selected_chem_id_dict[country])>0:\n",
    "            for chem in water_high_country_selected_chem_id_dict[country]:\n",
    "                try:\n",
    "                    z_list.append(real_significant_final_exp_proximity[chem,dis][1])\n",
    "                except:\n",
    "                    pass\n",
    "        dis_dict[dis]= np.round(np.mean(z_list),2)            \n",
    "    water_high_country_selected_dis_mean_prox_dict[country] = dis_dict\n",
    "\n",
    "water_high_country_selected_chem_id_gbd_dict = {}\n",
    "for country,dis_dict in water_high_country_selected_dis_mean_prox_dict.items():\n",
    "    if len(dis_dict)>0:\n",
    "        dis_z_list = []\n",
    "        for dis,z in dis_dict.items():\n",
    "            try:\n",
    "                dis_z_list.append([water_disgenet_gbd_dict[dis],z])\n",
    "            except:\n",
    "                pass\n",
    "        water_high_country_selected_chem_id_gbd_dict[country] = dis_z_list\n",
    "\n",
    "water_high_country_selected_chem_id_gbd_mean_dict = {}\n",
    "for country,dislist in water_high_country_selected_chem_id_gbd_dict.items():\n",
    "    dis_dict = {}\n",
    "    dis_list = []\n",
    "    for el in dislist:\n",
    "        dis_list.append(el[0])\n",
    "    dis_list_counter_dict = dict(Counter(dis_list).most_common())\n",
    "    for dis,count in dis_list_counter_dict.items():\n",
    "        z_list = []\n",
    "        if count>1:\n",
    "            for el in dislist:\n",
    "                if dis==el[0]:\n",
    "                    z_list.append(el[1])\n",
    "            dis_dict[dis]=np.round(np.mean(z_list),2)\n",
    "        else:\n",
    "            for el in dislist:\n",
    "                if dis==el[0]:\n",
    "                    dis_dict[el[0]] = el[1]\n",
    "    water_high_country_selected_chem_id_gbd_mean_dict[country]=dis_dict\n",
    "\n",
    "proximal_dis_list=[] \n",
    "for country, disdict in water_high_country_selected_chem_id_gbd_mean_dict.items():\n",
    "    proximal_dis_list.extend(list(disdict.keys()))\n",
    "    \n",
    "proximal_dis_incidence={}\n",
    "country_set=set(GBD_df['location'])\n",
    "for dis in set(proximal_dis_list):\n",
    "    incidence_list=[]\n",
    "    for country in country_set:\n",
    "        incidence_list.append(country_dis_incidence[country][dis])\n",
    "    proximal_dis_incidence[dis]=incidence_list\n",
    "\n",
    "\n",
    "proximal_dis_prevalence={}\n",
    "for dis in set(proximal_dis_list):\n",
    "    prevalence_list=[]\n",
    "    for country in country_set:\n",
    "        prevalence_list.append(country_dis_prevalence[country][dis])\n",
    "    proximal_dis_prevalence[dis]=prevalence_list\n",
    "\n",
    "proximal_dis_mortality={}\n",
    "for dis in set(proximal_dis_list):\n",
    "    mortality_list=[]\n",
    "    for country in country_set:\n",
    "        mortality_list.append(country_dis_mortality[country][dis])\n",
    "    proximal_dis_mortality[dis]=mortality_list\n",
    "\n",
    "proximal_dis_daly={}\n",
    "for dis in set(proximal_dis_list):\n",
    "    daly_list=[]\n",
    "    for country in country_set:\n",
    "        daly_list.append(country_dis_DALYs[country][dis])\n",
    "    proximal_dis_daly[dis]=daly_list\n",
    "\n",
    "\n",
    "water_high_country_proximal_zscore_dict={}\n",
    "for country,dis_dict in water_high_country_selected_chem_id_gbd_mean_dict.items():\n",
    "    country_proximal_zscore_list=[]\n",
    "    for dis in list(dis_dict.keys()):\n",
    "        mu = np.mean(proximal_dis_incidence[dis])\n",
    "        std = np.std(proximal_dis_incidence[dis])\n",
    "        z = (country_dis_incidence[country][dis]-mu)/std\n",
    "        country_proximal_zscore_list.append(z)\n",
    "        from scipy.stats import norm\n",
    "        ptemp = norm.sf(abs(z))\n",
    "    if len(country_proximal_zscore_list)>0:\n",
    "        water_high_country_proximal_zscore_dict[country]=country_proximal_zscore_list\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "9840690d-ac34-4f0f-89fd-84db00fcf104",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Distant\n",
    "high_country_water_faraway_dis_dict={}\n",
    "for country,chemset in water_high_country_selected_chem_id_dict.items():\n",
    "    if len(chemset)>0:\n",
    "        far_away_dis_set = []\n",
    "        for chem in chemset:\n",
    "            for chem_dis_pair in faraway_exp_dis_dict.keys():\n",
    "                if chem in chem_dis_pair[0]:\n",
    "                    far_away_dis_set.append(chem_dis_pair[1])\n",
    "        high_country_water_faraway_dis_dict[country] = set(far_away_dis_set)\n",
    "\n",
    "high_country_water_faraway_dis_dict = {}\n",
    "for country,dis_dict in water_high_country_selected_chem_id_gbd_mean_dict.items():\n",
    "    chem_set = water_high_country_selected_chem_id_dict[country]\n",
    "    country_faraway_dis_list = []\n",
    "    country_specific_faraway_dis_dict = {}\n",
    "    for dis in faraway_dis_list:\n",
    "        if dis not in dis_dict.keys():\n",
    "            country_faraway_dis_list.append(dis)\n",
    "    for dis in country_faraway_dis_list:\n",
    "        if dis not in faraway_manually_added_diseases:\n",
    "            disgenet_name_list = []\n",
    "            z_score_list = []\n",
    "            for disgenet_name,gbd_id in n_long_faraway_water_disset_gbd_dict_cleaned_v2.items():\n",
    "                if dis == gbd_id:\n",
    "                    disgenet_name_list.append(disgenet_name)\n",
    "                    for chem in chem_set:\n",
    "                        try:\n",
    "                            z_score_list.append(exp_proximity_not_significant[chem,disgenet_name][1])\n",
    "                        except:\n",
    "                            pass\n",
    "            if np.mean(z_score_list)>1.5:\n",
    "                country_specific_faraway_dis_dict[dis]=np.mean(z_score_list)\n",
    "        elif dis=='Congenital heart anomalies':\n",
    "            disgenet_name_list = ['Congenital Heart Defects','Deafness, Congenital Heart Defects, and Posterior Embryotoxon',\n",
    "                                 'Congenital long QT syndrome']\n",
    "            z_score_list = []\n",
    "            for disgenet_name in disgenet_name_list:\n",
    "                for chem in chem_set:\n",
    "                    try:\n",
    "                        z_score_list.append(exp_proximity_not_significant[chem,disgenet_name][1])\n",
    "                    except:\n",
    "                        pass\n",
    "            if np.mean(z_score_list)>1.5:\n",
    "                country_specific_faraway_dis_dict[dis]=np.mean(z_score_list)\n",
    "        \n",
    "        elif dis=='Congenital musculoskeletal and limb anomalies':\n",
    "            disgenet_name_list = ['Congenital Fiber Type Disproportion','Myopathy, Actin, Congenital, with Excess of Thin Myofilaments',\n",
    "                                 'Myopathy, Actin, Congenital, With Cores','Muscular Dystrophy, Congenital, Megaconial Type',\n",
    "                                 'Myotonia Congenita','Myasthenic Syndromes, Congenital','Muscular Dystrophy, Congenital, Megaconial Type',\n",
    "                                 'Congenital Myasthenic Syndromes, Postsynaptic','Congenital Myasthenic Syndromes, Presynaptic',\n",
    "                                 'Muscular Dystrophy, Congenital, Due To Integrin Alpha-7 Deficiency','Muscular Dystrophy, Congenital, Type 1D',\n",
    "                                 'Congenital Fibrosis of the Extraocular Muscles']\n",
    "            z_score_list = []\n",
    "            for disgenet_name in disgenet_name_list:\n",
    "                for chem in chem_set:\n",
    "                    try:\n",
    "                        z_score_list.append(exp_proximity_not_significant[chem,disgenet_name][1])\n",
    "                    except:\n",
    "                        pass\n",
    "            if np.mean(z_score_list)>1.5:\n",
    "                country_specific_faraway_dis_dict[dis]=np.mean(z_score_list)\n",
    "            \n",
    "        elif dis=='Congenital birth defects':\n",
    "            disgenet_name_list = [dis for dis in list(diseases_genes_associated_ppi.keys()) if 'Congenital' in dis or 'congenital' in dis]\n",
    "            z_score_list = []\n",
    "            for disgenet_name in disgenet_name_list:\n",
    "                for chem in chem_set:\n",
    "                    try:\n",
    "                        z_score_list.append(exp_proximity_not_significant[chem,disgenet_name][1])\n",
    "                    except:\n",
    "                        pass\n",
    "            if np.mean(z_score_list)>1.5:\n",
    "                country_specific_faraway_dis_dict[dis]=np.mean(z_score_list)\n",
    "\n",
    "        elif dis=='Turner syndrome':\n",
    "            disgenet_name_list = [dis for dis in list(diseases_genes_associated_ppi.keys()) if 'Turner' in dis]\n",
    "            z_score_list = []\n",
    "            for disgenet_name in disgenet_name_list:\n",
    "                for chem in chem_set:\n",
    "                    try:\n",
    "                        z_score_list.append(exp_proximity_not_significant[chem,disgenet_name][1])\n",
    "                    except:\n",
    "                        pass\n",
    "            if np.mean(z_score_list)>1.5:\n",
    "                country_specific_faraway_dis_dict[dis]=np.mean(z_score_list)\n",
    "    high_country_water_faraway_dis_dict[country] = country_specific_faraway_dis_dict\n",
    "\n",
    "\n",
    "high_country_noprox_water_dis_dict = {}\n",
    "for country,dis_dict in water_high_country_selected_chem_id_gbd_mean_dict.items():\n",
    "    n_dis = len(dis_dict)\n",
    "    chem_set = water_high_country_selected_chem_id_dict[country]\n",
    "    country_noprox_dis_list = []\n",
    "    country_specific_noprox_dis_dict = {}\n",
    "    \n",
    "    for gbd_dis,dislist in no_prox_air_disset_gbd_dict_cleaned_inv.items():\n",
    "        if gbd_dis not in dis_dict.keys():\n",
    "            if gbd_dis not in list(high_country_water_faraway_dis_dict[country].keys()):\n",
    "                country_noprox_dis_list.append(gbd_dis)\n",
    "    for dis in country_noprox_dis_list:\n",
    "        disgenet_name_list = no_prox_air_disset_gbd_dict_cleaned_inv[dis]\n",
    "        z_score_list = []\n",
    "        for chem in chem_set:\n",
    "            for disgenet_name in disgenet_name_list:\n",
    "                try:\n",
    "                    z_score_list.append(exp_proximity_not_significant[chem,disgenet_name][1])\n",
    "                except:\n",
    "                    pass\n",
    "        if np.mean(z_score_list)>-1.5:\n",
    "            country_specific_noprox_dis_dict[dis]=np.mean(z_score_list)\n",
    "    country_specific_noprox_dis_dict_sort = {k: v for k, v in sorted(country_specific_noprox_dis_dict.items(), key=lambda item: item[1])}\n",
    "    \n",
    "    if len(country_specific_noprox_dis_dict_sort)>n_dis:\n",
    "        country_specific_noprox_dis_dict_sort_keys = list(country_specific_noprox_dis_dict_sort.keys())\n",
    "        n_country_specific_noprox_dis_dict_sort = {}\n",
    "        for k in list(country_specific_noprox_dis_dict_sort_keys)[:n_dis]:\n",
    "            n_country_specific_noprox_dis_dict_sort[k]=country_specific_noprox_dis_dict_sort[k]\n",
    "        \n",
    "        high_country_noprox_water_dis_dict[country] = n_country_specific_noprox_dis_dict_sort\n",
    "    else:\n",
    "        high_country_noprox_water_dis_dict[country] = country_specific_noprox_dis_dict_sort\n",
    "\n",
    "#Let's just merge the no-proximal dictionary and the faraway dictionary\n",
    "new_high_water_country_distant_dis_dict = {}\n",
    "for country,dis_dict in water_high_country_selected_chem_id_gbd_mean_dict.items():\n",
    "    n_proxdis = len(dis_dict)\n",
    "    n_distantdis = len(high_country_water_faraway_dis_dict[country])\n",
    "    noprox_list = list(high_country_noprox_water_dis_dict[country].keys())\n",
    "    missing_dis = n_proxdis-n_distantdis\n",
    "    try:\n",
    "        dislist = noprox_list[:missing_dis]\n",
    "        distant_dict = high_country_water_faraway_dis_dict[country].copy()\n",
    "        for dis in dislist:\n",
    "            distant_dict[dis] = high_country_noprox_water_dis_dict[country][dis]\n",
    "        new_high_water_country_distant_dis_dict[country] = distant_dict\n",
    "    except:\n",
    "        distant_dict = {**high_country_water_faraway_dis_dict[country], **high_country_noprox_water_dis_dict[country]}\n",
    "        new_high_water_country_distant_dis_dict[country] = distant_dict\n",
    "\n",
    "distant_dis_list=[] \n",
    "for country, disdict in new_high_water_country_distant_dis_dict.items():\n",
    "    distant_dis_list.extend(list(disdict.keys()))\n",
    "    \n",
    "distant_dis_incidence={}\n",
    "country_set=set(GBD_df['location'])\n",
    "for dis in set(distant_dis_list):\n",
    "    incidence_list=[]\n",
    "    for country in country_set:\n",
    "        incidence_list.append(country_dis_incidence[country][dis])\n",
    "    distant_dis_incidence[dis]=incidence_list\n",
    "\n",
    "\n",
    "distant_dis_prevalence={}\n",
    "for dis in set(distant_dis_list):\n",
    "     prevalence_list=[]\n",
    "     for country in country_set:\n",
    "         prevalence_list.append(country_dis_prevalence[country][dis])\n",
    "     distant_dis_prevalence[dis]=prevalence_list\n",
    "\n",
    "distant_dis_mortality={}\n",
    "for dis in set(distant_dis_list):\n",
    "     mortality_list=[]\n",
    "     for country in country_set:\n",
    "         mortality_list.append(country_dis_mortality[country][dis])\n",
    "     distant_dis_mortality[dis]=mortality_list\n",
    "\n",
    "distant_dis_daly={}\n",
    "for dis in set(distant_dis_list):\n",
    "     daly_list=[]\n",
    "     for country in country_set:\n",
    "         daly_list.append(country_dis_DALYs[country][dis])\n",
    "     distant_dis_daly[dis]=daly_list\n",
    "\n",
    "\n",
    "water_high_country_distant_zscore_dict={}\n",
    "for country,dis_dict in new_high_water_country_distant_dis_dict.items():\n",
    "    country_distant_zscore_list=[]\n",
    "    for dis in list(dis_dict.keys()):\n",
    "        mu = np.mean(distant_dis_incidence[dis])\n",
    "        std = np.std(distant_dis_incidence[dis])\n",
    "        z = (country_dis_incidence[country][dis]-mu)/std\n",
    "        country_distant_zscore_list.append(z)\n",
    "\n",
    "        from scipy.stats import norm\n",
    "        ptemp = norm.sf(abs(z))\n",
    "    water_high_country_distant_zscore_dict[country]=country_distant_zscore_list"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "232d0d8d-6797-4247-89ea-f4cec999a962",
   "metadata": {},
   "outputs": [],
   "source": [
    "water_pollution_high_zscore_df = pd.DataFrame()\n",
    "\n",
    "selected_countries=set(water_high_country_proximal_zscore_dict.keys())\n",
    "tot_country_list=[]\n",
    "tot_zscore_list=[]\n",
    "location_network_list=[]\n",
    "for country in selected_countries:\n",
    "    farawaylist_cleaned = [x for x in water_high_country_distant_zscore_dict[country] if str(x) != 'nan']\n",
    "    proximallist_cleaned = [x for x in water_high_country_proximal_zscore_dict[country] if str(x) != 'nan']\n",
    "    tot_zscorelist_partial=farawaylist_cleaned+proximallist_cleaned\n",
    "    for i in proximallist_cleaned:\n",
    "        location_network_list.append(\"proximal\")\n",
    "        tot_zscore_list.append(i)\n",
    "        tot_country_list.append(country)\n",
    "    for i in farawaylist_cleaned:\n",
    "        location_network_list.append(\"distant\")\n",
    "        tot_zscore_list.append(i)\n",
    "        tot_country_list.append(country)\n",
    "\n",
    "water_pollution_high_zscore_df['location']=tot_country_list\n",
    "water_pollution_high_zscore_df['z-score']=tot_zscore_list\n",
    "water_pollution_high_zscore_df['network_location']=location_network_list\n",
    "\n",
    "my_order = sorted(list(set(water_pollution_high_zscore_df['location'])))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "e3d454a5-9286-4913-9bfb-4f10f016214e",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Let's test the difference between proximal and no-proximal (because with distant is kind of obvious and it doesn't\n",
    "#make so much sense)\n",
    "import scipy\n",
    "for country in water_high_country_proximal_zscore_dict.keys():\n",
    "    try:\n",
    "        proximallist_cleaned = [x for x in water_high_country_proximal_zscore_dict[country] if str(x) != 'nan']\n",
    "        distantlist_cleaned = [x for x in water_high_country_distant_zscore_dict[country] if str(x) != 'nan']\n",
    "        prox_mean = np.mean(proximallist_cleaned)\n",
    "        distant_mean = np.mean(distantlist_cleaned)\n",
    "        pval_noprox = scipy.stats.mannwhitneyu(proximallist_cleaned, distantlist_cleaned, use_continuity=True, alternative='greater')[1]\n",
    "        print(country,prox_mean,distant_mean,round(pval_noprox,4),round(pval_distant,4))\n",
    "    except:\n",
    "        print(country,prox_mean,distant_mean,round(pval_noprox,4))\n",
    "        "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "f786ab4a-ecce-4aaa-869d-a0786e0c8ea2",
   "metadata": {},
   "outputs": [],
   "source": [
    "import scipy \n",
    "# Assuming your data loading and plot setup code here\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(15, 20))\n",
    "sns.set(style=\"white\")\n",
    "g = sns.boxplot(x=\"z-score\", y=\"location\", orient=\"h\",\n",
    "                hue=\"network_location\", palette=[\"#2B7A98\", \"#54CAF2\"], order=my_order,\n",
    "                data=water_pollution_high_zscore_df)\n",
    "fontsize = 36\n",
    "plt.legend(loc=\"upper right\", frameon=True, fontsize=fontsize)\n",
    "plt.yticks(fontsize=36)\n",
    "\n",
    "for ax in plt.gcf().axes:\n",
    "    l = ax.get_xlabel()\n",
    "    ax.set_xlabel(l, fontsize=36)\n",
    "    ax.set_ylabel(l, fontsize=36)\n",
    "\n",
    "plt.ylabel('', fontsize=25)\n",
    "plt.xlabel('', fontsize=25)\n",
    "ax.spines['top'].set_visible(False)\n",
    "ax.spines['right'].set_visible(False)\n",
    "ax.spines['bottom'].set_visible(True)\n",
    "ax.spines['left'].set_visible(True)\n",
    "ax.set_xticklabels(ax.get_xticks(), size=36)\n",
    "\n",
    "pv_list = [] \n",
    "locations = [] \n",
    "stars = [] \n",
    "\n",
    "\n",
    "for country in my_order:\n",
    "    proximallist_cleaned = [x for x in water_high_country_proximal_zscore_dict[country] if str(x) != 'nan']\n",
    "    distantlist_cleaned = [x for x in water_high_country_distant_zscore_dict[country] if str(x) != 'nan']\n",
    "    prox_mean = np.mean(proximallist_cleaned)\n",
    "    distant_mean = np.mean(distantlist_cleaned)\n",
    "    pval_noprox = scipy.stats.mannwhitneyu(proximallist_cleaned, distantlist_cleaned, use_continuity=True, alternative='greater')[1]\n",
    "    pv_list.append(pval_noprox)\n",
    "    locations.append(country)\n",
    "    if pval_noprox<0.001:\n",
    "        stars.append(\"***\")\n",
    "    elif pval_noprox>0.001 and pval_noprox<0.01:\n",
    "        stars.append(\"**\")\n",
    "    elif pval_noprox>0.01 and pval_noprox<0.05:\n",
    "        stars.append(\"*\")\n",
    "    else:\n",
    "        stars.append(\"n.s.\")\n",
    "\n",
    "for pv, loc, star in zip(pv_list, locations, stars):\n",
    "    yloc = my_order.index(loc) + 0.4  \n",
    "    xloc = max(water_pollution_high_zscore_df[water_pollution_high_zscore_df['location'] == loc]['z-score']) + 0.5  \n",
    "  \n",
    "    # Add the annotation\n",
    "    ax.text(xloc, yloc, star, fontsize=fontsize, color=\"black\", ha='center')\n",
    "plt.savefig('output/Figures/SD11D_v01_post_rev.pdf',format='pdf',dpi=300, bbox_inches = \"tight\")\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "bee24bfe-49be-4c62-8b36-92dbb73ddacd",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "166ccfb6-311b-43f7-850e-c81a86da072f",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "5a2d6845-eff0-41a3-9f77-662dbb4411f7",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "5e5d99c9-f846-4bbc-88f3-96296c7f1019",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "202b9185-4f53-4094-8596-e5727fec769c",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Prevalence\n",
    "\n",
    "water_country_proximal_zscore_dict={}\n",
    "for country,dis_dict in water_high_country_selected_chem_id_gbd_mean_dict.items():\n",
    "    country_proximal_zscore_list=[]\n",
    "    for dis in list(dis_dict.keys()):\n",
    "        mu = np.mean(proximal_dis_prevalence[dis])\n",
    "        std = np.std(proximal_dis_prevalence[dis])\n",
    "        z = (country_dis_prevalence[country][dis]-mu)/std\n",
    "        country_proximal_zscore_list.append(z)\n",
    "        from scipy.stats import norm\n",
    "        ptemp = norm.sf(abs(z))\n",
    "    water_country_proximal_zscore_dict[country]=country_proximal_zscore_list\n",
    "\n",
    "\n",
    "water_country_distant_zscore_dict={}\n",
    "for country,dis_dict in new_high_water_country_distant_dis_dict.items():\n",
    "    country_distant_zscore_list=[]\n",
    "    for dis in list(dis_dict.keys()):\n",
    "        mu = np.mean(distant_dis_prevalence[dis])\n",
    "        std = np.std(distant_dis_prevalence[dis])\n",
    "        z = (country_dis_prevalence[country][dis]-mu)/std\n",
    "        country_distant_zscore_list.append(z)\n",
    "        from scipy.stats import norm\n",
    "        ptemp = norm.sf(abs(z))\n",
    "    water_country_distant_zscore_dict[country]=country_distant_zscore_list\n",
    "\n",
    "water_pollution_zscore_df = pd.DataFrame()\n",
    "selected_countries=set(water_country_distant_zscore_dict.keys())\n",
    "tot_country_list=[]\n",
    "tot_zscore_list=[]\n",
    "location_network_list=[]\n",
    "for country in selected_countries:\n",
    "    farawaylist_cleaned = [x for x in water_country_distant_zscore_dict[country] if str(x) != 'nan']\n",
    "    proximallist_cleaned = [x for x in water_country_proximal_zscore_dict[country] if str(x) != 'nan']\n",
    "    tot_zscorelist_partial=farawaylist_cleaned+proximallist_cleaned\n",
    "    for i in proximallist_cleaned:\n",
    "        location_network_list.append(\"proximal\")\n",
    "        tot_zscore_list.append(i)\n",
    "        tot_country_list.append(country)\n",
    "    for i in farawaylist_cleaned:\n",
    "        location_network_list.append(\"distant\")\n",
    "        tot_zscore_list.append(i)\n",
    "        tot_country_list.append(country)\n",
    "\n",
    "water_pollution_zscore_df['location']=tot_country_list\n",
    "water_pollution_zscore_df['z-score']=tot_zscore_list\n",
    "water_pollution_zscore_df['network_location']=location_network_list\n",
    "\n",
    "for country in water_country_proximal_zscore_dict.keys():\n",
    "    try:\n",
    "        proximallist_cleaned = [x for x in water_country_proximal_zscore_dict[country] if str(x) != 'nan']\n",
    "        distantlist_cleaned = [x for x in water_country_distant_zscore_dict[country] if str(x) != 'nan']\n",
    "        prox_mean = np.mean(proximallist_cleaned)\n",
    "        distant_mean = np.mean(distantlist_cleaned)\n",
    "        pval_noprox = scipy.stats.mannwhitneyu(proximallist_cleaned, distantlist_cleaned, use_continuity=True, alternative='greater')[1]\n",
    "        print(country,prox_mean,distant_mean,round(pval_noprox,4),round(pval_distant,4))\n",
    "    except:\n",
    "        print(country,prox_mean,distant_mean,round(pval_noprox,4))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "845ee8e5-2328-4220-b2c5-21bcfd39fed3",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "fig, ax = plt.subplots(figsize=(15, 20))\n",
    "sns.set(style=\"white\")\n",
    "g = sns.boxplot(x=\"z-score\", y=\"location\", orient=\"h\",\n",
    "                hue=\"network_location\", palette=[\"#2B7A98\", \"#54CAF2\"], order=my_order,\n",
    "                data=water_pollution_zscore_df)\n",
    "fontsize = 36\n",
    "plt.legend(loc=\"upper right\", frameon=True, fontsize=fontsize)\n",
    "plt.yticks(fontsize=36)\n",
    "\n",
    "for ax in plt.gcf().axes:\n",
    "    l = ax.get_xlabel()\n",
    "    ax.set_xlabel(l, fontsize=36)\n",
    "    ax.set_ylabel(l, fontsize=36)\n",
    "\n",
    "plt.ylabel('', fontsize=25)\n",
    "plt.xlabel('', fontsize=25)\n",
    "ax.spines['top'].set_visible(False)\n",
    "ax.spines['right'].set_visible(False)\n",
    "ax.spines['bottom'].set_visible(True)\n",
    "ax.spines['left'].set_visible(True)\n",
    "ax.set_xticklabels(ax.get_xticks(), size=36)\n",
    "\n",
    "pv_list = []  # Example p-values\n",
    "locations = []  # Example y-axis categories\n",
    "stars = []  # Based on the p-values, this is what we want to add\n",
    "\n",
    "\n",
    "\n",
    "for country in my_order:\n",
    "    proximallist_cleaned = [x for x in water_country_proximal_zscore_dict[country] if str(x) != 'nan']\n",
    "    distantlist_cleaned = [x for x in water_country_distant_zscore_dict[country] if str(x) != 'nan']\n",
    "    prox_mean = np.mean(proximallist_cleaned)\n",
    "    distant_mean = np.mean(distantlist_cleaned)\n",
    "    pval_noprox = scipy.stats.mannwhitneyu(proximallist_cleaned, distantlist_cleaned, use_continuity=True, alternative='greater')[1]\n",
    "    pv_list.append(pval_noprox)\n",
    "    locations.append(country)\n",
    "    if pval_noprox<0.001:\n",
    "        stars.append(\"***\")\n",
    "    elif pval_noprox>0.001 and pval_noprox<0.01:\n",
    "        stars.append(\"**\")\n",
    "    elif pval_noprox>0.01 and pval_noprox<0.05:\n",
    "        stars.append(\"*\")\n",
    "    else:\n",
    "        stars.append(\"n.s.\")\n",
    "\n",
    "\n",
    "for pv, loc, star in zip(pv_list, locations, stars):\n",
    "    yloc = my_order.index(loc) + 0.3  # This presumes 'my_order' list contains the order of y-categories\n",
    "    xloc = max(water_pollution_zscore_df[water_pollution_zscore_df['location'] == loc]['z-score']) + 0.5  \n",
    "    # Add the annotation\n",
    "    ax.text(xloc, yloc, star, fontsize=fontsize, color=\"black\", ha='center')\n",
    "\n",
    "plt.savefig('output/Figures/SD12F_v01_post_rev.pdf',format='pdf',dpi=300, bbox_inches = \"tight\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "591ec4cf-df1d-4543-84eb-00e84010f277",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "44fbf77d-4f20-4bec-913c-f120972f2221",
   "metadata": {},
   "outputs": [],
   "source": [
    "#mortality\n",
    "\n",
    "water_country_proximal_zscore_dict={}\n",
    "for country,dis_dict in water_high_country_selected_chem_id_gbd_mean_dict.items():\n",
    "    country_proximal_zscore_list=[]\n",
    "    for dis in list(dis_dict.keys()):\n",
    "        mu = np.mean(proximal_dis_mortality[dis])\n",
    "        std = np.std(proximal_dis_mortality[dis])\n",
    "        z = (country_dis_mortality[country][dis]-mu)/std\n",
    "        country_proximal_zscore_list.append(z)\n",
    "        from scipy.stats import norm\n",
    "        ptemp = norm.sf(abs(z))\n",
    "    water_country_proximal_zscore_dict[country]=country_proximal_zscore_list\n",
    "\n",
    "\n",
    "water_country_distant_zscore_dict={}\n",
    "for country,dis_dict in new_high_water_country_distant_dis_dict.items():\n",
    "    country_distant_zscore_list=[]\n",
    "    for dis in list(dis_dict.keys()):\n",
    "        mu = np.mean(distant_dis_mortality[dis])\n",
    "        std = np.std(distant_dis_mortality[dis])\n",
    "        z = (country_dis_mortality[country][dis]-mu)/std\n",
    "        country_distant_zscore_list.append(z)\n",
    "        from scipy.stats import norm\n",
    "        ptemp = norm.sf(abs(z))\n",
    "    water_country_distant_zscore_dict[country]=country_distant_zscore_list\n",
    "\n",
    "water_pollution_zscore_df = pd.DataFrame()\n",
    "selected_countries=set(water_country_distant_zscore_dict.keys())\n",
    "tot_country_list=[]\n",
    "tot_zscore_list=[]\n",
    "location_network_list=[]\n",
    "for country in selected_countries:\n",
    "    farawaylist_cleaned = [x for x in water_country_distant_zscore_dict[country] if str(x) != 'nan']\n",
    "    proximallist_cleaned = [x for x in water_country_proximal_zscore_dict[country] if str(x) != 'nan']\n",
    "    tot_zscorelist_partial=farawaylist_cleaned+proximallist_cleaned\n",
    "    for i in proximallist_cleaned:\n",
    "        location_network_list.append(\"proximal\")\n",
    "        tot_zscore_list.append(i)\n",
    "        tot_country_list.append(country)\n",
    "    for i in farawaylist_cleaned:\n",
    "        location_network_list.append(\"distant\")\n",
    "        tot_zscore_list.append(i)\n",
    "        tot_country_list.append(country)\n",
    "\n",
    "water_pollution_zscore_df['location']=tot_country_list\n",
    "water_pollution_zscore_df['z-score']=tot_zscore_list\n",
    "water_pollution_zscore_df['network_location']=location_network_list\n",
    "\n",
    "for country in water_country_proximal_zscore_dict.keys():\n",
    "    try:\n",
    "        proximallist_cleaned = [x for x in water_country_proximal_zscore_dict[country] if str(x) != 'nan']\n",
    "        distantlist_cleaned = [x for x in water_country_distant_zscore_dict[country] if str(x) != 'nan']\n",
    "        prox_mean = np.mean(proximallist_cleaned)\n",
    "        distant_mean = np.mean(distantlist_cleaned)\n",
    "        pval_noprox = scipy.stats.mannwhitneyu(proximallist_cleaned, distantlist_cleaned, use_continuity=True, alternative='greater')[1]\n",
    "        print(country,prox_mean,distant_mean,round(pval_noprox,4),round(pval_distant,4))\n",
    "    except:\n",
    "        print(country,prox_mean,distant_mean,round(pval_noprox,4))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "026cfac8-947c-446b-b009-28b24cb2a1a0",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "fig, ax = plt.subplots(figsize=(15, 20))\n",
    "sns.set(style=\"white\")\n",
    "g = sns.boxplot(x=\"z-score\", y=\"location\", orient=\"h\",\n",
    "                hue=\"network_location\", palette=[\"#2B7A98\", \"#54CAF2\"], order=my_order,\n",
    "                data=water_pollution_zscore_df)\n",
    "fontsize = 36\n",
    "plt.legend(loc=\"upper right\", frameon=True, fontsize=fontsize)\n",
    "plt.yticks(fontsize=36)\n",
    "\n",
    "for ax in plt.gcf().axes:\n",
    "    l = ax.get_xlabel()\n",
    "    ax.set_xlabel(l, fontsize=36)\n",
    "    ax.set_ylabel(l, fontsize=36)\n",
    "\n",
    "plt.ylabel('', fontsize=25)\n",
    "plt.xlabel('', fontsize=25)\n",
    "ax.spines['top'].set_visible(False)\n",
    "ax.spines['right'].set_visible(False)\n",
    "ax.spines['bottom'].set_visible(True)\n",
    "ax.spines['left'].set_visible(True)\n",
    "ax.set_xticklabels(ax.get_xticks(), size=36)\n",
    "\n",
    "pv_list = []  # Example p-values\n",
    "locations = []  # Example y-axis categories\n",
    "stars = []  # Based on the p-values, this is what we want to add\n",
    "\n",
    "\n",
    "\n",
    "for country in my_order:\n",
    "    proximallist_cleaned = [x for x in water_country_proximal_zscore_dict[country] if str(x) != 'nan']\n",
    "    distantlist_cleaned = [x for x in water_country_distant_zscore_dict[country] if str(x) != 'nan']\n",
    "    prox_mean = np.mean(proximallist_cleaned)\n",
    "    distant_mean = np.mean(distantlist_cleaned)\n",
    "    pval_noprox = scipy.stats.mannwhitneyu(proximallist_cleaned, distantlist_cleaned, use_continuity=True, alternative='greater')[1]\n",
    "    pv_list.append(pval_noprox)\n",
    "    locations.append(country)\n",
    "    if pval_noprox<0.001:\n",
    "        stars.append(\"***\")\n",
    "    elif pval_noprox>0.001 and pval_noprox<0.01:\n",
    "        stars.append(\"**\")\n",
    "    elif pval_noprox>0.01 and pval_noprox<0.05:\n",
    "        stars.append(\"*\")\n",
    "    else:\n",
    "        stars.append(\"n.s.\")\n",
    "\n",
    "\n",
    "for pv, loc, star in zip(pv_list, locations, stars):\n",
    "    yloc = my_order.index(loc) + 0.3  # This presumes 'my_order' list contains the order of y-categories\n",
    "    xloc = max(water_pollution_zscore_df[water_pollution_zscore_df['location'] == loc]['z-score']) + 0.5  \n",
    "    # Add the annotation\n",
    "    ax.text(xloc, yloc, star, fontsize=fontsize, color=\"black\", ha='center')\n",
    "\n",
    "plt.savefig('output/Figures/SD13F_v01_post_rev.pdf',format='pdf',dpi=300, bbox_inches = \"tight\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "6b428904-447f-43de-b6ea-97d35266422d",
   "metadata": {},
   "outputs": [],
   "source": [
    "#daly\n",
    "\n",
    "water_country_proximal_zscore_dict={}\n",
    "for country,dis_dict in water_high_country_selected_chem_id_gbd_mean_dict.items():\n",
    "    country_proximal_zscore_list=[]\n",
    "    for dis in list(dis_dict.keys()):\n",
    "        mu = np.mean(proximal_dis_daly[dis])\n",
    "        std = np.std(proximal_dis_daly[dis])\n",
    "        z = (country_dis_DALYs[country][dis]-mu)/std\n",
    "        country_proximal_zscore_list.append(z)\n",
    "        from scipy.stats import norm\n",
    "        ptemp = norm.sf(abs(z))\n",
    "    water_country_proximal_zscore_dict[country]=country_proximal_zscore_list\n",
    "\n",
    "\n",
    "water_country_distant_zscore_dict={}\n",
    "for country,dis_dict in new_high_water_country_distant_dis_dict.items():\n",
    "    country_distant_zscore_list=[]\n",
    "    for dis in list(dis_dict.keys()):\n",
    "        mu = np.mean(distant_dis_daly[dis])\n",
    "        std = np.std(distant_dis_daly[dis])\n",
    "        z = (country_dis_DALYs[country][dis]-mu)/std\n",
    "        country_distant_zscore_list.append(z)\n",
    "        from scipy.stats import norm\n",
    "        ptemp = norm.sf(abs(z))\n",
    "    water_country_distant_zscore_dict[country]=country_distant_zscore_list\n",
    "\n",
    "water_pollution_zscore_df = pd.DataFrame()\n",
    "selected_countries=set(water_country_distant_zscore_dict.keys())\n",
    "tot_country_list=[]\n",
    "tot_zscore_list=[]\n",
    "location_network_list=[]\n",
    "for country in selected_countries:\n",
    "    farawaylist_cleaned = [x for x in water_country_distant_zscore_dict[country] if str(x) != 'nan']\n",
    "    proximallist_cleaned = [x for x in water_country_proximal_zscore_dict[country] if str(x) != 'nan']\n",
    "    tot_zscorelist_partial=farawaylist_cleaned+proximallist_cleaned\n",
    "    for i in proximallist_cleaned:\n",
    "        location_network_list.append(\"proximal\")\n",
    "        tot_zscore_list.append(i)\n",
    "        tot_country_list.append(country)\n",
    "    for i in farawaylist_cleaned:\n",
    "        location_network_list.append(\"distant\")\n",
    "        tot_zscore_list.append(i)\n",
    "        tot_country_list.append(country)\n",
    "\n",
    "water_pollution_zscore_df['location']=tot_country_list\n",
    "water_pollution_zscore_df['z-score']=tot_zscore_list\n",
    "water_pollution_zscore_df['network_location']=location_network_list\n",
    "\n",
    "for country in water_country_proximal_zscore_dict.keys():\n",
    "    try:\n",
    "        proximallist_cleaned = [x for x in water_country_proximal_zscore_dict[country] if str(x) != 'nan']\n",
    "        distantlist_cleaned = [x for x in water_country_distant_zscore_dict[country] if str(x) != 'nan']\n",
    "        prox_mean = np.mean(proximallist_cleaned)\n",
    "        distant_mean = np.mean(distantlist_cleaned)\n",
    "        pval_noprox = scipy.stats.mannwhitneyu(proximallist_cleaned, distantlist_cleaned, use_continuity=True, alternative='greater')[1]\n",
    "        print(country,prox_mean,distant_mean,round(pval_noprox,4),round(pval_distant,4))\n",
    "    except:\n",
    "        print(country,prox_mean,distant_mean,round(pval_noprox,4))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "4b6e9351-c2a4-4bd0-bcbb-8fe8267d44cf",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "fig, ax = plt.subplots(figsize=(15, 20))\n",
    "sns.set(style=\"white\")\n",
    "g = sns.boxplot(x=\"z-score\", y=\"location\", orient=\"h\",\n",
    "                hue=\"network_location\", palette=[\"#2B7A98\", \"#54CAF2\"], order=my_order,\n",
    "                data=water_pollution_zscore_df)\n",
    "fontsize = 36\n",
    "plt.legend(loc=\"upper right\", frameon=True, fontsize=fontsize)\n",
    "plt.yticks(fontsize=36)\n",
    "\n",
    "for ax in plt.gcf().axes:\n",
    "    l = ax.get_xlabel()\n",
    "    ax.set_xlabel(l, fontsize=36)\n",
    "    ax.set_ylabel(l, fontsize=36)\n",
    "\n",
    "plt.ylabel('', fontsize=25)\n",
    "plt.xlabel('', fontsize=25)\n",
    "ax.spines['top'].set_visible(False)\n",
    "ax.spines['right'].set_visible(False)\n",
    "ax.spines['bottom'].set_visible(True)\n",
    "ax.spines['left'].set_visible(True)\n",
    "ax.set_xticklabels(ax.get_xticks(), size=36)\n",
    "\n",
    "pv_list = []  # Example p-values\n",
    "locations = []  # Example y-axis categories\n",
    "stars = []  # Based on the p-values, this is what we want to add\n",
    "\n",
    "\n",
    "\n",
    "for country in my_order:\n",
    "    proximallist_cleaned = [x for x in water_country_proximal_zscore_dict[country] if str(x) != 'nan']\n",
    "    distantlist_cleaned = [x for x in water_country_distant_zscore_dict[country] if str(x) != 'nan']\n",
    "    prox_mean = np.mean(proximallist_cleaned)\n",
    "    distant_mean = np.mean(distantlist_cleaned)\n",
    "    pval_noprox = scipy.stats.mannwhitneyu(proximallist_cleaned, distantlist_cleaned, use_continuity=True, alternative='greater')[1]\n",
    "    pv_list.append(pval_noprox)\n",
    "    locations.append(country)\n",
    "    if pval_noprox<0.001:\n",
    "        stars.append(\"***\")\n",
    "    elif pval_noprox>0.001 and pval_noprox<0.01:\n",
    "        stars.append(\"**\")\n",
    "    elif pval_noprox>0.01 and pval_noprox<0.05:\n",
    "        stars.append(\"*\")\n",
    "    else:\n",
    "        stars.append(\"n.s.\")\n",
    "\n",
    "\n",
    "for pv, loc, star in zip(pv_list, locations, stars):\n",
    "    yloc = my_order.index(loc) + 0.3  # This presumes 'my_order' list contains the order of y-categories\n",
    "    xloc = max(water_pollution_zscore_df[water_pollution_zscore_df['location'] == loc]['z-score']) + 0.5  \n",
    "    # Add the annotation\n",
    "    ax.text(xloc, yloc, star, fontsize=fontsize, color=\"black\", ha='center')\n",
    "\n",
    "plt.savefig('output/Figures/SD14F_v01_post_rev.pdf',format='pdf',dpi=300, bbox_inches = \"tight\")\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "b0b572a2-8bf0-4125-95ed-11f5fc20c2db",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "940380a4-f2f3-4d64-8b81-95c7b730df3b",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "32c0ce2e-a10e-4e79-8c49-e29810fad757",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "9f9f6232-4c43-4802-a43d-8a14fbd61e93",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d4520f1f-450c-4b7a-a540-2a28c78cde00",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "2b545820-de07-40ab-98ee-c95c50f7837e",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "a3310a9e-0194-4c9a-9a5d-a5ce2d16aa1d",
   "metadata": {},
   "outputs": [],
   "source": [
    "GBD_df_ti = GBD_df[GBD_df['year']==2000]\n",
    "GBD_df_tf = GBD_df[GBD_df['year']==2019]\n",
    "\n",
    "filtered_df = GBD_df_ti[GBD_df_ti['measure'] == 'Incidence']\n",
    "\n",
    "country_dis_incidence_ti = {}\n",
    "\n",
    "for country in filtered_df['location'].unique():\n",
    "    country_df = filtered_df[filtered_df['location'] == country].set_index('cause')['val']\n",
    "    country_dis_incidence_ti[country] = country_df.to_dict()\n",
    "\n",
    "country_dis_incidence_tf = {}\n",
    "filtered_df = GBD_df_tf[GBD_df_tf['measure'] == 'Incidence']\n",
    "for country in filtered_df['location'].unique():\n",
    "    country_df = filtered_df[filtered_df['location'] == country].set_index('cause')['val']\n",
    "    country_dis_incidence_tf[country] = country_df.to_dict()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "54ae7aeb-75a5-4c94-80fa-dc1a7bb19566",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "4dd39a8d-33c9-407c-b758-c96fb0ed3e58",
   "metadata": {},
   "outputs": [],
   "source": [
    "proximal_dis_list=[] \n",
    "for country, disdict in country_selected_chem_id_gbd_mean_dict.items():\n",
    "    proximal_dis_list.extend(list(disdict.keys()))\n",
    "    \n",
    "proximal_dis_incidence_ti={}\n",
    "country_set=set(GBD_df['location'])\n",
    "for dis in set(proximal_dis_list):\n",
    "    incidence_list=[]\n",
    "    for country in country_set:\n",
    "        try:\n",
    "            incidence_list.append(country_dis_incidence_ti[country][dis])\n",
    "        except:\n",
    "            pass\n",
    "    proximal_dis_incidence_ti[dis]=incidence_list\n",
    "\n",
    "country_proximal_zscore_dict_ti={}\n",
    "for country,dis_dict in country_selected_chem_id_gbd_mean_dict.items():\n",
    "    country_proximal_zscore_list=[]\n",
    "    for dis in list(dis_dict.keys()):\n",
    "        if dis in country_dis_incidence_ti[country].keys():\n",
    "            mu = np.mean(proximal_dis_incidence_ti[dis])\n",
    "            std = np.std(proximal_dis_incidence_ti[dis])\n",
    "            z = (country_dis_incidence_ti[country][dis]-mu)/std\n",
    "            country_proximal_zscore_list.append(z)\n",
    "            from scipy.stats import norm\n",
    "            ptemp = norm.sf(abs(z))\n",
    "    country_proximal_zscore_dict_ti[country]=country_proximal_zscore_list\n",
    "\n",
    "proximal_dis_incidence_tf={}\n",
    "for dis in set(proximal_dis_list):\n",
    "    incidence_list=[]\n",
    "    for country in country_set:\n",
    "        try:\n",
    "            incidence_list.append(country_dis_incidence_tf[country][dis])\n",
    "        except:\n",
    "            pass\n",
    "    proximal_dis_incidence_tf[dis]=incidence_list\n",
    "\n",
    "country_proximal_zscore_dict_tf={}\n",
    "for country,dis_dict in country_selected_chem_id_gbd_mean_dict.items():\n",
    "    country_proximal_zscore_list=[]\n",
    "    for dis in list(dis_dict.keys()):\n",
    "        if dis in country_dis_incidence_tf[country].keys():\n",
    "            mu = np.mean(proximal_dis_incidence_tf[dis])\n",
    "            std = np.std(proximal_dis_incidence_tf[dis])\n",
    "            z = (country_dis_incidence_tf[country][dis]-mu)/std\n",
    "            country_proximal_zscore_list.append(z)\n",
    "            from scipy.stats import norm\n",
    "            ptemp = norm.sf(abs(z))\n",
    "    country_proximal_zscore_dict_tf[country]=country_proximal_zscore_list\n",
    "\n",
    "air_temporal_proximal_zscore_df = pd.DataFrame()\n",
    "\n",
    "selected_countries=set(country_proximal_zscore_dict_tf.keys())\n",
    "tot_country_list=[]\n",
    "tot_zscore_list=[]\n",
    "time_list=[]\n",
    "for country in selected_countries:\n",
    "    tf_cleaned = [x for x in country_proximal_zscore_dict_tf[country] if str(x) != 'nan']\n",
    "    ti_cleaned = [x for x in country_proximal_zscore_dict_ti[country] if str(x) != 'nan']\n",
    "    tot_zscorelist_partial=tf_cleaned+ti_cleaned\n",
    "    for i in tf_cleaned:\n",
    "        time_list.append(\"final time\")\n",
    "        tot_zscore_list.append(i)\n",
    "        tot_country_list.append(country)\n",
    "    for i in ti_cleaned:\n",
    "        time_list.append(\"initial time\")\n",
    "        tot_zscore_list.append(i)\n",
    "        tot_country_list.append(country)\n",
    "\n",
    "air_temporal_proximal_zscore_df['location']=tot_country_list\n",
    "air_temporal_proximal_zscore_df['z-score']=tot_zscore_list\n",
    "air_temporal_proximal_zscore_df['time']=time_list\n",
    "\n",
    "my_order = sorted(list(set(air_temporal_proximal_zscore_df['location'])))\n",
    "\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(15, 20))\n",
    "sns.set(style=\"white\")\n",
    "g = sns.boxplot(x=\"z-score\", y=\"location\", orient=\"h\",\n",
    "                hue=\"time\", palette=[\"r\", \"b\"], order=my_order,\n",
    "                data=air_temporal_proximal_zscore_df)\n",
    "fontsize = 36\n",
    "plt.legend(loc=\"upper right\", frameon=True, fontsize=fontsize)\n",
    "plt.yticks(fontsize=36)\n",
    "\n",
    "for ax in plt.gcf().axes:\n",
    "    l = ax.get_xlabel()\n",
    "    ax.set_xlabel(l, fontsize=36)\n",
    "    ax.set_ylabel(l, fontsize=36)\n",
    "\n",
    "plt.ylabel('', fontsize=25)\n",
    "plt.xlabel('', fontsize=25)\n",
    "ax.spines['top'].set_visible(False)\n",
    "ax.spines['right'].set_visible(False)\n",
    "ax.spines['bottom'].set_visible(True)\n",
    "ax.spines['left'].set_visible(True)\n",
    "ax.set_xticklabels(ax.get_xticks(), size=36)\n",
    "\n",
    "pv_list = []  \n",
    "locations = []  \n",
    "stars = []  \n",
    "\n",
    "for country in my_order:\n",
    "    tf_cleaned = [x for x in country_proximal_zscore_dict_tf[country] if str(x) != 'nan']\n",
    "    ti_cleaned = [x for x in country_proximal_zscore_dict_ti[country] if str(x) != 'nan']\n",
    "    tf_mean = np.mean(tf_cleaned)\n",
    "    ti_mean = np.mean(ti_cleaned)\n",
    "    pval_noprox = scipy.stats.mannwhitneyu(tf_cleaned, ti_cleaned, use_continuity=True, alternative='greater')[1]\n",
    "    pv_list.append(pval_noprox)\n",
    "    locations.append(country)\n",
    "    if pval_noprox<0.001:\n",
    "        stars.append(\"***\")\n",
    "    elif pval_noprox>0.001 and pval_noprox<0.01:\n",
    "        stars.append(\"**\")\n",
    "    elif pval_noprox>0.01 and pval_noprox<0.05:\n",
    "        stars.append(\"*\")\n",
    "    else:\n",
    "        stars.append(\"n.s.\")\n",
    "\n",
    "for pv, loc, star in zip(pv_list, locations, stars):\n",
    "    yloc = my_order.index(loc) + 0.4 \n",
    "    xloc = max(air_temporal_proximal_zscore_df[air_temporal_proximal_zscore_df['location'] == loc]['z-score']) + 0.5  \n",
    "    ax.text(xloc, yloc, star, fontsize=fontsize, color=\"black\", ha='center')\n",
    "\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "f8dfa238-0651-4a87-826b-513a8065327a",
   "metadata": {},
   "outputs": [],
   "source": [
    "filtered_df = GBD_df_ti[GBD_df_ti['measure'] == 'Prevalence']\n",
    "\n",
    "country_dis_prevalence_ti = {}\n",
    "\n",
    "for country in filtered_df['location'].unique():\n",
    "    country_df = filtered_df[filtered_df['location'] == country].set_index('cause')['val']\n",
    "    country_dis_prevalence_ti[country] = country_df.to_dict()\n",
    "\n",
    "country_dis_prevalence_tf = {}\n",
    "filtered_df = GBD_df_tf[GBD_df_tf['measure'] == 'Prevalence']\n",
    "for country in filtered_df['location'].unique():\n",
    "    country_df = filtered_df[filtered_df['location'] == country].set_index('cause')['val']\n",
    "    country_dis_prevalence_tf[country] = country_df.to_dict()\n",
    "\n",
    "proximal_dis_list=[] \n",
    "for country, disdict in country_selected_chem_id_gbd_mean_dict.items():\n",
    "    proximal_dis_list.extend(list(disdict.keys()))\n",
    "    \n",
    "proximal_dis_prevalence_ti={}\n",
    "country_set=set(GBD_df['location'])\n",
    "for dis in set(proximal_dis_list):\n",
    "    prevalence_list=[]\n",
    "    for country in country_set:\n",
    "        try:\n",
    "            prevalence_list.append(country_dis_prevalence_ti[country][dis])\n",
    "        except:\n",
    "            pass\n",
    "    proximal_dis_prevalence_ti[dis]=prevalence_list\n",
    "\n",
    "country_proximal_zscore_dict_ti={}\n",
    "for country,dis_dict in country_selected_chem_id_gbd_mean_dict.items():\n",
    "    country_proximal_zscore_list=[]\n",
    "    for dis in list(dis_dict.keys()):\n",
    "        if dis in country_dis_prevalence_ti[country].keys():\n",
    "            mu = np.mean(proximal_dis_prevalence_ti[dis])\n",
    "            std = np.std(proximal_dis_prevalence_ti[dis])\n",
    "            z = (country_dis_prevalence_ti[country][dis]-mu)/std\n",
    "            country_proximal_zscore_list.append(z)\n",
    "            from scipy.stats import norm\n",
    "            ptemp = norm.sf(abs(z))\n",
    "    country_proximal_zscore_dict_ti[country]=country_proximal_zscore_list\n",
    "\n",
    "proximal_dis_prevalence_tf={}\n",
    "for dis in set(proximal_dis_list):\n",
    "    prevalence_list=[]\n",
    "    for country in country_set:\n",
    "        try:\n",
    "            prevalence_list.append(country_dis_prevalence_tf[country][dis])\n",
    "        except:\n",
    "            pass\n",
    "    proximal_dis_prevalence_tf[dis]=prevalence_list\n",
    "\n",
    "country_proximal_zscore_dict_tf={}\n",
    "for country,dis_dict in country_selected_chem_id_gbd_mean_dict.items():\n",
    "    country_proximal_zscore_list=[]\n",
    "    for dis in list(dis_dict.keys()):\n",
    "        if dis in country_dis_prevalence_tf[country].keys():\n",
    "            mu = np.mean(proximal_dis_prevalence_tf[dis])\n",
    "            std = np.std(proximal_dis_prevalence_tf[dis])\n",
    "            z = (country_dis_prevalence_tf[country][dis]-mu)/std\n",
    "            country_proximal_zscore_list.append(z)\n",
    "            from scipy.stats import norm\n",
    "            ptemp = norm.sf(abs(z))\n",
    "    country_proximal_zscore_dict_tf[country]=country_proximal_zscore_list\n",
    "\n",
    "air_temporal_proximal_zscore_df = pd.DataFrame()\n",
    "\n",
    "selected_countries=set(country_proximal_zscore_dict_tf.keys())\n",
    "tot_country_list=[]\n",
    "tot_zscore_list=[]\n",
    "time_list=[]\n",
    "for country in selected_countries:\n",
    "    tf_cleaned = [x for x in country_proximal_zscore_dict_tf[country] if str(x) != 'nan']\n",
    "    ti_cleaned = [x for x in country_proximal_zscore_dict_ti[country] if str(x) != 'nan']\n",
    "    tot_zscorelist_partial=tf_cleaned+ti_cleaned\n",
    "    for i in tf_cleaned:\n",
    "        time_list.append(\"final time\")\n",
    "        tot_zscore_list.append(i)\n",
    "        tot_country_list.append(country)\n",
    "    for i in ti_cleaned:\n",
    "        time_list.append(\"initial time\")\n",
    "        tot_zscore_list.append(i)\n",
    "        tot_country_list.append(country)\n",
    "\n",
    "air_temporal_proximal_zscore_df['location']=tot_country_list\n",
    "air_temporal_proximal_zscore_df['z-score']=tot_zscore_list\n",
    "air_temporal_proximal_zscore_df['time']=time_list\n",
    "\n",
    "my_order = sorted(list(set(air_temporal_proximal_zscore_df['location'])))\n",
    "\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(15, 20))\n",
    "sns.set(style=\"white\")\n",
    "g = sns.boxplot(x=\"z-score\", y=\"location\", orient=\"h\",\n",
    "                hue=\"time\", palette=[\"r\", \"b\"], order=my_order,\n",
    "                data=air_temporal_proximal_zscore_df)\n",
    "fontsize = 36\n",
    "plt.legend(loc=\"upper right\", frameon=True, fontsize=fontsize)\n",
    "plt.yticks(fontsize=36)\n",
    "\n",
    "for ax in plt.gcf().axes:\n",
    "    l = ax.get_xlabel()\n",
    "    ax.set_xlabel(l, fontsize=36)\n",
    "    ax.set_ylabel(l, fontsize=36)\n",
    "\n",
    "plt.ylabel('', fontsize=25)\n",
    "plt.xlabel('', fontsize=25)\n",
    "ax.spines['top'].set_visible(False)\n",
    "ax.spines['right'].set_visible(False)\n",
    "ax.spines['bottom'].set_visible(True)\n",
    "ax.spines['left'].set_visible(True)\n",
    "ax.set_xticklabels(ax.get_xticks(), size=36)\n",
    "\n",
    "pv_list = []  \n",
    "locations = []  \n",
    "stars = []  \n",
    "\n",
    "for country in my_order:\n",
    "    tf_cleaned = [x for x in country_proximal_zscore_dict_tf[country] if str(x) != 'nan']\n",
    "    ti_cleaned = [x for x in country_proximal_zscore_dict_ti[country] if str(x) != 'nan']\n",
    "    tf_mean = np.mean(tf_cleaned)\n",
    "    ti_mean = np.mean(ti_cleaned)\n",
    "    pval_noprox = scipy.stats.mannwhitneyu(tf_cleaned, ti_cleaned, use_continuity=True, alternative='greater')[1]\n",
    "    pv_list.append(pval_noprox)\n",
    "    locations.append(country)\n",
    "    if pval_noprox<0.001:\n",
    "        stars.append(\"***\")\n",
    "    elif pval_noprox>0.001 and pval_noprox<0.01:\n",
    "        stars.append(\"**\")\n",
    "    elif pval_noprox>0.01 and pval_noprox<0.05:\n",
    "        stars.append(\"*\")\n",
    "    else:\n",
    "        stars.append(\"n.s.\")\n",
    "\n",
    "for pv, loc, star in zip(pv_list, locations, stars):\n",
    "    yloc = my_order.index(loc) + 0.4 \n",
    "    xloc = max(air_temporal_proximal_zscore_df[air_temporal_proximal_zscore_df['location'] == loc]['z-score']) + 0.5  \n",
    "    ax.text(xloc, yloc, star, fontsize=fontsize, color=\"black\", ha='center')\n",
    "\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "254a5d79-881e-4a2b-b303-ffe32d76815a",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "da5e2c3e-e9cd-4b71-aa2d-07a485b17c85",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ff82e938-057e-4b6d-900a-45a8d880c779",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Let's select those chemicals that have low concentrations at ti but high at tf"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "43b0dcb4-29c6-4585-8268-06f215ef707a",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "country_increasing_chemicals = {}\n",
    "\n",
    "grouped = who_air_nation.groupby('countryName')\n",
    "\n",
    "# Iterate over each country group\n",
    "for country, group in grouped:\n",
    "    increasing_pollutants = []\n",
    "    for _, row in group.iterrows():\n",
    "        pollutant = row['pollutant']\n",
    "        if row[2019] > row[2007]:  \n",
    "            increasing_pollutants.append(pollutant)\n",
    "    country_increasing_chemicals[country] = increasing_pollutants\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "6d7058e4-8d22-4e5d-8d6f-9f364df9a66d",
   "metadata": {},
   "outputs": [],
   "source": [
    "country_increasing_chemicals_id_dict = {}\n",
    "\n",
    "chem_id_list = list(chem_id_name_conversion.keys())\n",
    "\n",
    "for country,chem_list in country_increasing_chemicals.items():\n",
    "    id_chem_list=[]\n",
    "    for chem in chem_list:\n",
    "        try:\n",
    "            if chem!='PCDD + PCDF (dioxins + furans) (as Teq)':   \n",
    "                if air_chem_name_conversion_dict[chem] in chem_id_list:\n",
    "                    id_chem_list.append(air_chem_name_conversion_dict[chem])\n",
    "            else:\n",
    "                if 'D005663' in chem_id_list:\n",
    "                    id_chem_list.append('D005663')\n",
    "                if 'D000072316' in chem_id_list:\n",
    "                    id_chem_list.append('D000072316')\n",
    "                if 'D004147' in chem_id_list:\n",
    "                    id_chem_list.append('D004147')\n",
    "        except:\n",
    "            pass\n",
    "    if len(set(id_chem_list))>0:\n",
    "        country_increasing_chemicals_id_dict[country]=set(id_chem_list)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "8afb16a5-6a4f-443c-8b48-79f6fe6b4e61",
   "metadata": {},
   "outputs": [],
   "source": [
    "country_selected_dis_dict={}\n",
    "for country, chem_list in country_increasing_chemicals_id_dict.items():\n",
    "    dis_list=[]\n",
    "    for chem in chem_list:\n",
    "        for chem_dis_pair,val in real_significant_final_exp_proximity.items():\n",
    "            if chem in chem_dis_pair[0]:\n",
    "                dis_list.append(chem_dis_pair[1])\n",
    "    country_selected_dis_dict[country]=set(dis_list)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "9bc364ba-d4c2-4fa7-8915-eb44be6d16e4",
   "metadata": {},
   "outputs": [],
   "source": [
    "country_selected_dis_mean_prox_dict = {}\n",
    "for country,disset in country_selected_dis_dict.items():\n",
    "    dis_dict = {}\n",
    "    for dis in disset:\n",
    "        z_list = [] \n",
    "        if len(country_increasing_chemicals_id_dict[country])>0:\n",
    "            for chem in country_increasing_chemicals_id_dict[country]:\n",
    "                try:\n",
    "                    z_list.append(real_significant_final_exp_proximity[chem,dis][1])\n",
    "                except:\n",
    "                    pass\n",
    "        dis_dict[dis]= np.round(np.mean(z_list),2)\n",
    "                    \n",
    "    country_selected_dis_mean_prox_dict[country] = dis_dict\n",
    "\n",
    "#Here, we have to clean the diseases that appears in each countries, considering only the ones that are reported in\n",
    "#GBD and using a little of manual curation for merging some terms (for example breast cancer and carcinoma of breast)\n",
    "#Now, we can create a country proximal diseases dictionary with the diseases taken from GBD\n",
    "\n",
    "\n",
    "country_selected_chem_id_gbd_dict = {}\n",
    "for country,dis_dict in country_selected_dis_mean_prox_dict.items():\n",
    "    if len(dis_dict)>0:\n",
    "        dis_z_list = []\n",
    "        for dis,z in dis_dict.items():\n",
    "            try:\n",
    "                dis_z_list.append([air_disgenet_gbd_dict[dis],z])\n",
    "            except:\n",
    "                pass\n",
    "        country_selected_chem_id_gbd_dict[country] = dis_z_list\n",
    "\n",
    "\n",
    "\n",
    "country_selected_chem_id_gbd_mean_dict = {}\n",
    "for country,dislist in country_selected_chem_id_gbd_dict.items():\n",
    "    dis_dict = {}\n",
    "    dis_list = []\n",
    "    for el in dislist:\n",
    "        dis_list.append(el[0])\n",
    "    dis_list_counter_dict = dict(Counter(dis_list).most_common())\n",
    "    for dis,count in dis_list_counter_dict.items():\n",
    "        z_list = []\n",
    "        if count>1:\n",
    "            for el in dislist:\n",
    "                if dis==el[0]:\n",
    "                    z_list.append(el[1])\n",
    "            dis_dict[dis]=np.round(np.mean(z_list),2)\n",
    "        else:\n",
    "            for el in dislist:\n",
    "                if dis==el[0]:\n",
    "                    dis_dict[el[0]] = el[1]\n",
    "    country_selected_chem_id_gbd_mean_dict[country]=dis_dict"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "f91dad6e-b3cb-42ad-8b38-78e4ab949ec0",
   "metadata": {},
   "outputs": [],
   "source": [
    "GBD_df_ti = GBD_df[GBD_df['year']==2007]\n",
    "GBD_df_tf = GBD_df[GBD_df['year']==2019]\n",
    "\n",
    "filtered_df = GBD_df_ti[GBD_df_ti['measure'] == 'Incidence']\n",
    "\n",
    "country_dis_incidence_ti = {}\n",
    "\n",
    "for country in filtered_df['location'].unique():\n",
    "    country_df = filtered_df[filtered_df['location'] == country].set_index('cause')['val']\n",
    "    country_dis_incidence_ti[country] = country_df.to_dict()\n",
    "\n",
    "country_dis_incidence_tf = {}\n",
    "filtered_df = GBD_df_tf[GBD_df_tf['measure'] == 'Incidence']\n",
    "for country in filtered_df['location'].unique():\n",
    "    country_df = filtered_df[filtered_df['location'] == country].set_index('cause')['val']\n",
    "    country_dis_incidence_tf[country] = country_df.to_dict()\n",
    "    \n",
    "proximal_dis_list=[] \n",
    "for country, disdict in country_selected_chem_id_gbd_mean_dict.items():\n",
    "    proximal_dis_list.extend(list(disdict.keys()))\n",
    "    \n",
    "proximal_dis_incidence_ti={}\n",
    "country_set=set(GBD_df['location'])\n",
    "for dis in set(proximal_dis_list):\n",
    "    incidence_list=[]\n",
    "    for country in country_set:\n",
    "        try:\n",
    "            incidence_list.append(country_dis_incidence_ti[country][dis])\n",
    "        except:\n",
    "            pass\n",
    "    proximal_dis_incidence_ti[dis]=incidence_list\n",
    "\n",
    "country_proximal_zscore_dict_ti={}\n",
    "for country,dis_dict in country_selected_chem_id_gbd_mean_dict.items():\n",
    "    country_proximal_zscore_list=[]\n",
    "    for dis in list(dis_dict.keys()):\n",
    "        if dis in country_dis_incidence_ti[country].keys():\n",
    "            mu = np.mean(proximal_dis_incidence_ti[dis])\n",
    "            std = np.std(proximal_dis_incidence_ti[dis])\n",
    "            z = (country_dis_incidence_ti[country][dis]-mu)/std\n",
    "            country_proximal_zscore_list.append(z)\n",
    "            from scipy.stats import norm\n",
    "            ptemp = norm.sf(abs(z))\n",
    "    country_proximal_zscore_dict_ti[country]=country_proximal_zscore_list\n",
    "\n",
    "proximal_dis_incidence_tf={}\n",
    "for dis in set(proximal_dis_list):\n",
    "    incidence_list=[]\n",
    "    for country in country_set:\n",
    "        try:\n",
    "            incidence_list.append(country_dis_incidence_tf[country][dis])\n",
    "        except:\n",
    "            pass\n",
    "    proximal_dis_incidence_tf[dis]=incidence_list\n",
    "\n",
    "country_proximal_zscore_dict_tf={}\n",
    "for country,dis_dict in country_selected_chem_id_gbd_mean_dict.items():\n",
    "    country_proximal_zscore_list=[]\n",
    "    for dis in list(dis_dict.keys()):\n",
    "        if dis in country_dis_incidence_tf[country].keys():\n",
    "            mu = np.mean(proximal_dis_incidence_tf[dis])\n",
    "            std = np.std(proximal_dis_incidence_tf[dis])\n",
    "            z = (country_dis_incidence_tf[country][dis]-mu)/std\n",
    "            country_proximal_zscore_list.append(z)\n",
    "            from scipy.stats import norm\n",
    "            ptemp = norm.sf(abs(z))\n",
    "    country_proximal_zscore_dict_tf[country]=country_proximal_zscore_list\n",
    "\n",
    "air_temporal_proximal_zscore_df = pd.DataFrame()\n",
    "\n",
    "selected_countries=set(country_proximal_zscore_dict_tf.keys())\n",
    "tot_country_list=[]\n",
    "tot_zscore_list=[]\n",
    "time_list=[]\n",
    "for country in selected_countries:\n",
    "    tf_cleaned = [x for x in country_proximal_zscore_dict_tf[country] if str(x) != 'nan']\n",
    "    ti_cleaned = [x for x in country_proximal_zscore_dict_ti[country] if str(x) != 'nan']\n",
    "    tot_zscorelist_partial=tf_cleaned+ti_cleaned\n",
    "    for i in tf_cleaned:\n",
    "        time_list.append(\"final time\")\n",
    "        tot_zscore_list.append(i)\n",
    "        tot_country_list.append(country)\n",
    "    for i in ti_cleaned:\n",
    "        time_list.append(\"initial time\")\n",
    "        tot_zscore_list.append(i)\n",
    "        tot_country_list.append(country)\n",
    "\n",
    "air_temporal_proximal_zscore_df['location']=tot_country_list\n",
    "air_temporal_proximal_zscore_df['z-score']=tot_zscore_list\n",
    "air_temporal_proximal_zscore_df['time']=time_list\n",
    "\n",
    "my_order = sorted(list(set(air_temporal_proximal_zscore_df['location'])))\n",
    "\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(15, 20))\n",
    "sns.set(style=\"white\")\n",
    "g = sns.boxplot(x=\"z-score\", y=\"location\", orient=\"h\",\n",
    "                hue=\"time\", palette=[\"r\", \"b\"], order=my_order,\n",
    "                data=air_temporal_proximal_zscore_df)\n",
    "fontsize = 36\n",
    "plt.legend(loc=\"upper right\", frameon=True, fontsize=fontsize)\n",
    "plt.yticks(fontsize=36)\n",
    "\n",
    "for ax in plt.gcf().axes:\n",
    "    l = ax.get_xlabel()\n",
    "    ax.set_xlabel(l, fontsize=36)\n",
    "    ax.set_ylabel(l, fontsize=36)\n",
    "\n",
    "plt.ylabel('', fontsize=25)\n",
    "plt.xlabel('', fontsize=25)\n",
    "ax.spines['top'].set_visible(False)\n",
    "ax.spines['right'].set_visible(False)\n",
    "ax.spines['bottom'].set_visible(True)\n",
    "ax.spines['left'].set_visible(True)\n",
    "ax.set_xticklabels(ax.get_xticks(), size=36)\n",
    "\n",
    "pv_list = []  \n",
    "locations = []  \n",
    "stars = []  \n",
    "\n",
    "for country in my_order:\n",
    "    tf_cleaned = [x for x in country_proximal_zscore_dict_tf[country] if str(x) != 'nan']\n",
    "    ti_cleaned = [x for x in country_proximal_zscore_dict_ti[country] if str(x) != 'nan']\n",
    "    tf_mean = np.mean(tf_cleaned)\n",
    "    ti_mean = np.mean(ti_cleaned)\n",
    "    pval_noprox = scipy.stats.mannwhitneyu(tf_cleaned, ti_cleaned, use_continuity=True, alternative='greater')[1]\n",
    "    pv_list.append(pval_noprox)\n",
    "    locations.append(country)\n",
    "    if pval_noprox<0.001:\n",
    "        stars.append(\"***\")\n",
    "    elif pval_noprox>0.001 and pval_noprox<0.01:\n",
    "        stars.append(\"**\")\n",
    "    elif pval_noprox>0.01 and pval_noprox<0.05:\n",
    "        stars.append(\"*\")\n",
    "    else:\n",
    "        stars.append(\"n.s.\")\n",
    "\n",
    "for pv, loc, star in zip(pv_list, locations, stars):\n",
    "    yloc = my_order.index(loc) + 0.4 \n",
    "    xloc = max(air_temporal_proximal_zscore_df[air_temporal_proximal_zscore_df['location'] == loc]['z-score']) + 0.5  \n",
    "    ax.text(xloc, yloc, star, fontsize=fontsize, color=\"black\", ha='center')\n",
    "\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c75df72a-646f-42d3-a0a3-b4995d492d42",
   "metadata": {},
   "outputs": [],
   "source": [
    "GBD_df_ti = GBD_df[GBD_df['year']==2007]\n",
    "GBD_df_tf = GBD_df[GBD_df['year']==2019]\n",
    "\n",
    "filtered_df = GBD_df_ti[GBD_df_ti['measure'] == 'Prevalence']\n",
    "\n",
    "country_dis_prevalence_ti = {}\n",
    "\n",
    "for country in filtered_df['location'].unique():\n",
    "    country_df = filtered_df[filtered_df['location'] == country].set_index('cause')['val']\n",
    "    country_dis_prevalence_ti[country] = country_df.to_dict()\n",
    "\n",
    "country_dis_prevalence_tf = {}\n",
    "filtered_df = GBD_df_tf[GBD_df_tf['measure'] == 'Prevalence']\n",
    "for country in filtered_df['location'].unique():\n",
    "    country_df = filtered_df[filtered_df['location'] == country].set_index('cause')['val']\n",
    "    country_dis_prevalence_tf[country] = country_df.to_dict()\n",
    "    \n",
    "proximal_dis_list=[] \n",
    "for country, disdict in country_selected_chem_id_gbd_mean_dict.items():\n",
    "    proximal_dis_list.extend(list(disdict.keys()))\n",
    "    \n",
    "proximal_dis_prevalence_ti={}\n",
    "country_set=set(GBD_df['location'])\n",
    "for dis in set(proximal_dis_list):\n",
    "    prevalence_list=[]\n",
    "    for country in country_set:\n",
    "        try:\n",
    "            prevalence_list.append(country_dis_prevalence_ti[country][dis])\n",
    "        except:\n",
    "            pass\n",
    "    proximal_dis_prevalence_ti[dis]=prevalence_list\n",
    "\n",
    "country_proximal_zscore_dict_ti={}\n",
    "for country,dis_dict in country_selected_chem_id_gbd_mean_dict.items():\n",
    "    country_proximal_zscore_list=[]\n",
    "    for dis in list(dis_dict.keys()):\n",
    "        if dis in country_dis_prevalence_ti[country].keys():\n",
    "            mu = np.mean(proximal_dis_prevalence_ti[dis])\n",
    "            std = np.std(proximal_dis_prevalence_ti[dis])\n",
    "            z = (country_dis_prevalence_ti[country][dis]-mu)/std\n",
    "            country_proximal_zscore_list.append(z)\n",
    "            from scipy.stats import norm\n",
    "            ptemp = norm.sf(abs(z))\n",
    "    country_proximal_zscore_dict_ti[country]=country_proximal_zscore_list\n",
    "\n",
    "proximal_dis_prevalence_tf={}\n",
    "for dis in set(proximal_dis_list):\n",
    "    prevalence_list=[]\n",
    "    for country in country_set:\n",
    "        try:\n",
    "            prevalence_list.append(country_dis_prevalence_tf[country][dis])\n",
    "        except:\n",
    "            pass\n",
    "    proximal_dis_prevalence_tf[dis]=prevalence_list\n",
    "\n",
    "country_proximal_zscore_dict_tf={}\n",
    "for country,dis_dict in country_selected_chem_id_gbd_mean_dict.items():\n",
    "    country_proximal_zscore_list=[]\n",
    "    for dis in list(dis_dict.keys()):\n",
    "        if dis in country_dis_prevalence_tf[country].keys():\n",
    "            mu = np.mean(proximal_dis_prevalence_tf[dis])\n",
    "            std = np.std(proximal_dis_prevalence_tf[dis])\n",
    "            z = (country_dis_prevalence_tf[country][dis]-mu)/std\n",
    "            country_proximal_zscore_list.append(z)\n",
    "            from scipy.stats import norm\n",
    "            ptemp = norm.sf(abs(z))\n",
    "    country_proximal_zscore_dict_tf[country]=country_proximal_zscore_list\n",
    "\n",
    "air_temporal_proximal_zscore_df = pd.DataFrame()\n",
    "\n",
    "selected_countries=set(country_proximal_zscore_dict_tf.keys())\n",
    "tot_country_list=[]\n",
    "tot_zscore_list=[]\n",
    "time_list=[]\n",
    "for country in selected_countries:\n",
    "    tf_cleaned = [x for x in country_proximal_zscore_dict_tf[country] if str(x) != 'nan']\n",
    "    ti_cleaned = [x for x in country_proximal_zscore_dict_ti[country] if str(x) != 'nan']\n",
    "    tot_zscorelist_partial=tf_cleaned+ti_cleaned\n",
    "    for i in tf_cleaned:\n",
    "        time_list.append(\"final time\")\n",
    "        tot_zscore_list.append(i)\n",
    "        tot_country_list.append(country)\n",
    "    for i in ti_cleaned:\n",
    "        time_list.append(\"initial time\")\n",
    "        tot_zscore_list.append(i)\n",
    "        tot_country_list.append(country)\n",
    "\n",
    "air_temporal_proximal_zscore_df['location']=tot_country_list\n",
    "air_temporal_proximal_zscore_df['z-score']=tot_zscore_list\n",
    "air_temporal_proximal_zscore_df['time']=time_list\n",
    "\n",
    "my_order = sorted(list(set(air_temporal_proximal_zscore_df['location'])))\n",
    "\n",
    "\n",
    "fig, ax = plt.subplots(figsize=(15, 20))\n",
    "sns.set(style=\"white\")\n",
    "g = sns.boxplot(x=\"z-score\", y=\"location\", orient=\"h\",\n",
    "                hue=\"time\", palette=[\"r\", \"b\"], order=my_order,\n",
    "                data=air_temporal_proximal_zscore_df)\n",
    "fontsize = 36\n",
    "plt.legend(loc=\"upper right\", frameon=True, fontsize=fontsize)\n",
    "plt.yticks(fontsize=36)\n",
    "\n",
    "for ax in plt.gcf().axes:\n",
    "    l = ax.get_xlabel()\n",
    "    ax.set_xlabel(l, fontsize=36)\n",
    "    ax.set_ylabel(l, fontsize=36)\n",
    "\n",
    "plt.ylabel('', fontsize=25)\n",
    "plt.xlabel('', fontsize=25)\n",
    "ax.spines['top'].set_visible(False)\n",
    "ax.spines['right'].set_visible(False)\n",
    "ax.spines['bottom'].set_visible(True)\n",
    "ax.spines['left'].set_visible(True)\n",
    "ax.set_xticklabels(ax.get_xticks(), size=36)\n",
    "\n",
    "pv_list = []  \n",
    "locations = []  \n",
    "stars = []  \n",
    "\n",
    "for country in my_order:\n",
    "    tf_cleaned = [x for x in country_proximal_zscore_dict_tf[country] if str(x) != 'nan']\n",
    "    ti_cleaned = [x for x in country_proximal_zscore_dict_ti[country] if str(x) != 'nan']\n",
    "    tf_mean = np.mean(tf_cleaned)\n",
    "    ti_mean = np.mean(ti_cleaned)\n",
    "    pval_noprox = scipy.stats.mannwhitneyu(tf_cleaned, ti_cleaned, use_continuity=True, alternative='greater')[1]\n",
    "    pv_list.append(pval_noprox)\n",
    "    locations.append(country)\n",
    "    if pval_noprox<0.001:\n",
    "        stars.append(\"***\")\n",
    "    elif pval_noprox>0.001 and pval_noprox<0.01:\n",
    "        stars.append(\"**\")\n",
    "    elif pval_noprox>0.01 and pval_noprox<0.05:\n",
    "        stars.append(\"*\")\n",
    "    else:\n",
    "        stars.append(\"n.s.\")\n",
    "\n",
    "for pv, loc, star in zip(pv_list, locations, stars):\n",
    "    yloc = my_order.index(loc) + 0.4 \n",
    "    xloc = max(air_temporal_proximal_zscore_df[air_temporal_proximal_zscore_df['location'] == loc]['z-score']) + 0.5  \n",
    "    ax.text(xloc, yloc, star, fontsize=fontsize, color=\"black\", ha='center')\n",
    "\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "dcb17330-239d-4d71-9152-3facc3bd9e83",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "2ca6d1a7-30c7-4270-9b1d-f2f00942ce42",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "a30fff78-fa05-483c-8bcf-854967336422",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "94b2ad64-78c2-4844-bb57-a68153971ba8",
   "metadata": {},
   "outputs": [],
   "source": [
    "distant_dis_list=[] \n",
    "for country, disdict in new_country_distant_dis_dict.items():\n",
    "    distant_dis_list.extend(list(disdict.keys()))\n",
    "    \n",
    "distant_dis_incidence={}\n",
    "country_set=set(GBD_df['location'])\n",
    "for dis in set(distant_dis_list):\n",
    "    incidence_list=[]\n",
    "    for country in country_set:\n",
    "        incidence_list.append(country_dis_incidence[country][dis])\n",
    "    distant_dis_incidence[dis]=incidence_list\n",
    "\n",
    "distant_dis_prevalence={}\n",
    "for dis in set(distant_dis_list):\n",
    "    prevalence_list=[]\n",
    "    for country in country_set:\n",
    "        prevalence_list.append(country_dis_prevalence[country][dis])\n",
    "    distant_dis_prevalence[dis]=prevalence_list\n",
    "\n",
    "distant_dis_mortality={}\n",
    "for dis in set(distant_dis_list):\n",
    "    mortality_list=[]\n",
    "    for country in country_set:\n",
    "        mortality_list.append(country_dis_mortality[country][dis])\n",
    "    distant_dis_mortality[dis]=mortality_list\n",
    "\n",
    "distant_dis_daly={}\n",
    "for dis in set(distant_dis_list):\n",
    "    daly_list=[]\n",
    "    for country in country_set:\n",
    "        daly_list.append(country_dis_DALYs[country][dis])\n",
    "    distant_dis_daly[dis]=daly_list"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "58a201ee-e92c-40e3-8ff1-a2c9abec0860",
   "metadata": {},
   "outputs": [],
   "source": [
    "air_country_distant_zscore_dict={}\n",
    "\n",
    "for country,dis_dict in new_country_distant_dis_dict.items():\n",
    "    country_distant_zscore_list=[]\n",
    "    for dis in list(dis_dict.keys()):\n",
    "        mu = np.mean(distant_dis_incidence[dis])\n",
    "        std = np.std(distant_dis_incidence[dis])\n",
    "\n",
    "        z = (country_dis_incidence[country][dis]-mu)/std\n",
    "        country_distant_zscore_list.append(z)\n",
    "\n",
    "        from scipy.stats import norm\n",
    "        ptemp = norm.sf(abs(z))\n",
    "    air_country_distant_zscore_dict[country]=country_distant_zscore_list"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "365ffe3f-6ae7-408d-8270-aa9e36343a3c",
   "metadata": {},
   "outputs": [],
   "source": [
    "air_pollution_zscore_df = pd.DataFrame()\n",
    "\n",
    "selected_countries=set(air_country_distant_zscore_dict.keys())\n",
    "tot_country_list=[]\n",
    "tot_zscore_list=[]\n",
    "location_network_list=[]\n",
    "for country in selected_countries:\n",
    "    farawaylist_cleaned = [x for x in air_country_distant_zscore_dict[country] if str(x) != 'nan']\n",
    "    proximallist_cleaned = [x for x in country_proximal_zscore_dict[country] if str(x) != 'nan']\n",
    "    tot_zscorelist_partial=farawaylist_cleaned+proximallist_cleaned\n",
    "    for i in proximallist_cleaned:\n",
    "        location_network_list.append(\"proximal\")\n",
    "        tot_zscore_list.append(i)\n",
    "        tot_country_list.append(country)\n",
    "    for i in farawaylist_cleaned:\n",
    "        location_network_list.append(\"distant\")\n",
    "        tot_zscore_list.append(i)\n",
    "        tot_country_list.append(country)\n",
    "\n",
    "air_pollution_zscore_df['location']=tot_country_list\n",
    "air_pollution_zscore_df['z-score']=tot_zscore_list\n",
    "air_pollution_zscore_df['network_location']=location_network_list"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "8a9f0ad7-e993-4c5f-86ec-6561135ed1f3",
   "metadata": {},
   "outputs": [],
   "source": [
    "my_order = sorted(list(set(air_pollution_zscore_df['location'])))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "a3c22484-92a4-45bf-9f69-58891ec66e60",
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "\n"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.11.10"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}