BRAFProtocol.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "961898db",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import xml.etree.ElementTree as ET\n",
    "import os\n",
    "from urllib.request import urlretrieve as download\n",
    "from glob import glob as g\n",
    "from Bio.Blast import NCBIWWW, NCBIXML\n",
    "from mypdb import PDB_file as mypdb\n",
    "from Bio.Blast.Applications import NcbipsiblastCommandline\n",
    "from time import time as t\n",
    "from tqdm import tqdm\n",
    "\n",
    "\n",
    "import glob\n",
    "import xml.etree.ElementTree as ET\n",
    "from collections import defaultdict\n",
    "\n",
    "import re\n",
    "import MDAnalysis as mda\n",
    "from time import time as t"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ffd8d4e7",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Class to handle and download the hits from a blast search\n",
    "\n",
    "\n",
    "\n",
    "def merge_dicts(dict1, dict2):\n",
    "    for k2, v2 in dict2.items():\n",
    "        if k2 in dict1:\n",
    "            if isinstance(dict1[k2], list):\n",
    "                dict1[k2].append(v2)\n",
    "            else:\n",
    "                dict1[k2] = [dict1[k2], v2]\n",
    "        else:\n",
    "            print(f\"New Key : {k2}\")\n",
    "            dict1[k2] = v2\n",
    "    return dict1\n",
    "\n",
    "class hit:\n",
    "    def __init__(self, xml_hit):\n",
    "        self.hit_num = xml_hit[0].text\n",
    "        pdb_info = xml_hit[1].text.split(\"|\")[1:]\n",
    "        self.pdb_id = pdb_info[0]\n",
    "        self.pdb_chain = pdb_info[1]\n",
    "        self.description = xml_hit[2].text\n",
    "        data = [[y.tag.replace(\"Hsp_\", \"\"), y.text] for y in xml_hit[5][0]]\n",
    "        self.data = dict(data)\n",
    "        self.pdb = None\n",
    "        self._not_exists = False\n",
    "\n",
    "    @staticmethod\n",
    "    def download2(code, pdir=None):\n",
    "        base_url = \"https://files.rcsb.org/download\"\n",
    "        pdb_url = f\"{base_url}/{code}.pdb\"\n",
    "        f_p = os.path.join(pdir, f\"{code}.pdb\")\n",
    "        try:\n",
    "            download(pdb_url, f_p)\n",
    "            return f_p  # Return the file path if succeeded\n",
    "        except Exception:\n",
    "            print(f\"File {code} not found.\")\n",
    "            return None  # Return None if failed\n",
    "\n",
    "    def _assign(self, f_p):\n",
    "        self.pdb_path = f_p\n",
    "        self.pdb = mypdb(f_p)\n",
    "        \n",
    "    def check_for_pdb(self, pdir=None):\n",
    "        if pdir is None:\n",
    "            pdir = \"./pdbs/\"\n",
    "        elif pdir[-1] != \"/\":\n",
    "            pdir += \"/\"\n",
    "        if not os.path.isdir(pdir):\n",
    "            os.makedirs(pdir, exist_ok=True)\n",
    "        matches = g(pdir + f\"{self.pdb_id}*\")\n",
    "        if matches:\n",
    "            if not self.pdb:\n",
    "                self._assign(matches[0])\n",
    "            return True\n",
    "        else:\n",
    "            return False"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "ab2139fe",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Class to handle the blast search\n",
    "class blast:\n",
    "    def __init__(self, sequence, name, database, path, program=\"blastp\", hitlen=10000):\n",
    "        self.sequence = sequence\n",
    "        self.program = program\n",
    "        self.database = database\n",
    "        self.path = path\n",
    "        self.name = name\n",
    "        self.hitlen = hitlen\n",
    "\n",
    "        if program == \"blastp\" and (self.path and os.path.isfile(self.path)):\n",
    "            self.parse_search()\n",
    "        elif program == \"psiblast\":\n",
    "            self.psiblast_search()\n",
    "        else:\n",
    "            self.bsearch()\n",
    "\n",
    "    def bsearch(self):\n",
    "        # Perform the initial BLASTP search\n",
    "        if self.program == \"blastp\":\n",
    "            print(\"Searching BLASTP...\")\n",
    "            t1 = t()\n",
    "            self.results = NCBIWWW.qblast(self.program, self.database, self.sequence, hitlist_size=self.hitlen)\n",
    "            t2 = t()\n",
    "            print(f\"BLASTP took {round(t2-t1,4)} seconds\")\n",
    "            if not self.path:\n",
    "                self.path = f\"{self.name}-blast.xml\"\n",
    "            with open(self.path, \"w\") as output_xml:\n",
    "                output_xml.write(self.results.read())\n",
    "            self.parse_search()\n",
    "\n",
    "    def psiblast_search(self):\n",
    "        # Run local PSI-BLAST using NcbipsiblastCommandline\n",
    "        print(f\"Running PSI-BLAST on {self.name}...\")\n",
    "        input_fasta = f\"{self.name}.fasta\"\n",
    "        with open(input_fasta, \"w\") as f:\n",
    "            f.write(f\">query\\n{self.sequence}\\n\")\n",
    "\n",
    "        psiblast_cline = NcbipsiblastCommandline(\n",
    "            query=input_fasta,\n",
    "            db=\"/home/marmatt/ncbi-blast-2.16.0+/bin/pdbaa\",\n",
    "            evalue=10,\n",
    "            num_iterations=3,\n",
    "            out_ascii_pssm=f\"{self.name}.pssm\",\n",
    "            out=f\"{self.name}-psiblast.xml\",\n",
    "            outfmt=5\n",
    "        )\n",
    "        stdout, stderr = psiblast_cline()\n",
    "        if stderr:\n",
    "            print(f\"PSI-BLAST ERROR: {stderr}\")\n",
    "        else:\n",
    "            print(\"PSI-BLAST search completed.\")\n",
    "            self.parse_search(xml_file=f\"{self.name}-psiblast.xml\")\n",
    "\n",
    "    def parse_search(self, xml_file=None):\n",
    "        xml_file = xml_file or self.path\n",
    "        if not xml_file:\n",
    "            raise Exception(\"No XML file path provided.\")\n",
    "        t1 = t()\n",
    "        tree = ET.parse(xml_file)\n",
    "        iteration = tree.findall(\"./BlastOutput_iterations/Iteration/\")\n",
    "        self.query_length = iteration[3].text\n",
    "        hits = [hit(x) for x in iteration[-2]] #returns hit objects\n",
    "        self.hits = hits\n",
    "        mega_dict = hits[0].data\n",
    "        for x in hits[1:]:\n",
    "            mega_dict = merge_dicts(mega_dict, x.data)\n",
    "        mega_dict[\"PDB ID\"] = [x.pdb_id for x in hits]\n",
    "        mega_dict[\"Chain\"] = [x.pdb_chain for x in hits]\n",
    "        mega_dict[\"Description\"] = [x.description for x in hits]\n",
    "        self.df = pd.DataFrame.from_dict(mega_dict)\n",
    "        print(self.df)\n",
    "        t2 = t()\n",
    "        print(f\"Time taken to parse {t2-t1}\")\n",
    "    \n",
    "    def download_pdbs(self, pdir=None):\n",
    "        default_dir = \"./PDBs\"\n",
    "        pdir = os.path.abspath(pdir if pdir else default_dir)\n",
    "        if not os.path.isdir(pdir):\n",
    "            os.makedirs(pdir, exist_ok=True)\n",
    "        \n",
    "        files = [os.path.splitext(f)[0] for f in os.listdir(pdir)]\n",
    "        hit_bar = tqdm(self.hits, desc=\"Processing Hits\")\n",
    "        \n",
    "        for x in hit_bar:\n",
    "            if x.pdb_id not in files:\n",
    "                hit_bar.set_description(f\"Downloading {x}\")\n",
    "                \n",
    "                try:\n",
    "                    file_path = x.download2(x.pdb_id, pdir=pdir)\n",
    "\n",
    "                    if file_path:\n",
    "                        x._assign(file_path)\n",
    "                    else:\n",
    "                        # In case download2 does not return a valid path\n",
    "                        raise Exception(\"Download failed\")\n",
    "                except Exception as e:\n",
    "                    # Print a message if download fails or file path is invalid\n",
    "                    print(f\"Structure {x.pdb_id} was not found...\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e81f2b43",
   "metadata": {},
   "source": [
    "## Actually running the blast search"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "125b6d45",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "\"\"\"\n",
    "Here we perform a blast search on:\n",
    " 1.   The BRAF monomer\n",
    "\"\"\"\n",
    "with open(\"./fastas.txt\") as f:\n",
    "    fastas = f.readlines()\n",
    "\n",
    "braf_fasta = fastas[1]\n",
    "name = \"braf\"\n",
    "\n",
    "# Directory for storing blast search results\n",
    "output_dir = \"./blast_search\"\n",
    "# Ensure the directory exists\n",
    "os.makedirs(output_dir, exist_ok=True)\n",
    "\n",
    "res_path = os.path.join(output_dir, f\"{name}-blast.xml\")  # Path for the results file\n",
    "\n",
    "# Check for saved results in the specified directory\n",
    "if os.path.exists(res_path):\n",
    "    bs = blast(braf_fasta, name, database=\"pdb\", path=res_path)\n",
    "else:\n",
    "    bs = blast(braf_fasta, name, database=\"pdb\", path=None)\n",
    "\n",
    "bs.download_pdbs()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3b84c34d",
   "metadata": {},
   "source": [
    "## Counting the number of pdb files in the directory"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "2c13cfea",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "\n",
    "import glob\n",
    "def count_pdb_files(directory):\n",
    "    # Ensure the directory path ends with a slash\n",
    "    directory = os.path.join(directory, '')\n",
    "\n",
    "    # Use glob to find all .pdb files in the directory\n",
    "    pdb_files = glob.glob(os.path.join(directory, '*.pdb'))\n",
    "\n",
    "    # Return the count of .pdb files\n",
    "    return len(pdb_files)\n",
    "\n",
    "# Specify the directory\n",
    "pdb_directory = 'PDBs'\n",
    "\n",
    "# Get the count of PDB files\n",
    "pdb_count = count_pdb_files(pdb_directory)\n",
    "\n",
    "print(f\"There are {pdb_count} PDB files in the directory '{pdb_directory}'.\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "af58070b",
   "metadata": {},
   "source": [
    "## Counting the number of pdb hits in the xml file and checking which ones have not been downloaded\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "c0522d85",
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "\n",
    "def count_pdb_files(directory):\n",
    "    # Use glob to find all .pdb files in the directory\n",
    "    pdb_files = glob.glob(os.path.join(directory, '*.pdb'))\n",
    "\n",
    "    # Extract the base filenames (without extension) to compare with PDB IDs\n",
    "    pdb_file_ids = {os.path.splitext(os.path.basename(f))[0] for f in pdb_files}\n",
    "\n",
    "    return pdb_file_ids\n",
    "\n",
    "def find_unique_and_duplicate_pdb_hit_ids(xml_file):\n",
    "    # Parse the XML file\n",
    "    tree = ET.parse(xml_file)\n",
    "    root = tree.getroot()\n",
    "\n",
    "    # Dictionary to count occurrences of each PDB hit ID\n",
    "    hit_id_counts = defaultdict(int)\n",
    "\n",
    "    # Iterate over all Hit elements in the XML\n",
    "    for hit in root.findall('.//Hit'):\n",
    "        # Extract the Hit_id text\n",
    "        hit_id = hit.find('Hit_id').text\n",
    "\n",
    "        # Assuming the Hit_id format is 'pdb|PDB_ID|Chain', extract the PDB_ID\n",
    "        pdb_id = hit_id.split('|')[1]\n",
    "\n",
    "        # Increment the count for this PDB_ID\n",
    "        hit_id_counts[pdb_id] += 1\n",
    "\n",
    "    # Find all PDB IDs (unique and duplicates)\n",
    "    all_hit_ids = set(hit_id_counts.keys())\n",
    "\n",
    "    return all_hit_ids\n",
    "\n",
    "# Specify the directory and XML file\n",
    "pdb_directory = 'PDBs'\n",
    "xml_file = 'braf-blast.xml'\n",
    "\n",
    "# Get the PDB file IDs from the directory\n",
    "pdb_file_ids = count_pdb_files(pdb_directory)\n",
    "\n",
    "# Get all PDB hit IDs from the XML\n",
    "all_pdb_ids = find_unique_and_duplicate_pdb_hit_ids(xml_file)\n",
    "\n",
    "# Calculate the number of total PDB hits\n",
    "total_pdb_hits = len(all_pdb_ids)\n",
    "\n",
    "# Find PDB IDs in XML that are not in the directory\n",
    "missing_pdb_ids = all_pdb_ids - pdb_file_ids\n",
    "\n",
    "print(f\"There are {total_pdb_hits} total PDB hits in the file '{xml_file}'.\")\n",
    "print(f\"There are {len(missing_pdb_ids)} PDB IDs in the XML not found in the directory '{pdb_directory}':\")\n",
    "print(missing_pdb_ids)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "952b9660",
   "metadata": {},
   "source": [
    "## Here we are stripping the downloaded pdb files to only contain the chain of interest\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "4fc269f2",
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "import os\n",
    "from glob import glob\n",
    "import re\n",
    "import MDAnalysis as mda\n",
    "from time import time as t\n",
    "!ls\n",
    "def sglob(fp, absolute=True):\n",
    "    fps = sorted(glob(fp))\n",
    "    if absolute:\n",
    "        fps = [os.path.abspath(f) for f in fps]\n",
    "    return fps\n",
    "\n",
    "def strip_to_chain(pdb_file, chain_ID):\n",
    "    u = mda.Universe(pdb_file)\n",
    "    print(f\"Loaded trajectory from {pdb_file} with {len(u.atoms)} atoms.\")\n",
    "\n",
    "    chain = u.select_atoms(f\"protein and chainID {chain_ID}\")\n",
    "    if len(chain) == 0:\n",
    "        print(f\"Chain {chain_ID} not found in {pdb_file}.\")\n",
    "        return None\n",
    "    return chain\n",
    "\n",
    "def post_process(fname):\n",
    "    with open(fname, \"r\") as f_o:\n",
    "        initial_lines = f_o.readlines()\n",
    "\n",
    "    print(f\"File {fname} before post_process, first few lines:\")\n",
    "    print(\"\".join(initial_lines[:20]))\n",
    "\n",
    "    final_lines = initial_lines[-2:].copy()\n",
    "    no_ter = [line for line in initial_lines if line[:3] != \"TER\" or line in final_lines]\n",
    "\n",
    "    if len(no_ter) != len(initial_lines):\n",
    "        with open(fname, \"w\") as f_o:\n",
    "            print(f\"Rewriting {fname}, lines reduced from {len(initial_lines)} to {len(no_ter)}\")\n",
    "            f_o.write(\"\".join(no_ter))\n",
    "\n",
    "    print(f\"File {fname} after post_process, first few lines:\")\n",
    "    with open(fname, \"r\") as f_r:\n",
    "        print(\"\".join(f_r.readlines()[:20]))\n",
    "\n",
    "def parse_xml(xml_file):\n",
    "    hit_id = re.compile(r\"<Hit_id>(.*?)<.Hit_id>\")\n",
    "    with open(xml_file, \"r\") as f:\n",
    "        text = f.read()\n",
    "        results = [h.split(\"|\")[1:] for h in hit_id.findall(text)]\n",
    "        pdb_chain_dict = {}\n",
    "        for r in results:\n",
    "            pdb_chain_dict[r[0]+f\"_{r[1]}\"] = r[1]\n",
    "    return pdb_chain_dict\n",
    "\n",
    "def get_pdb_id(fp):\n",
    "    fp = fp.rsplit(\".\", 1)[0]\n",
    "    if \"/\" in fp:\n",
    "        fp = fp.rsplit(\"/\", 1)[1]\n",
    "    return fp\n",
    "\n",
    "def target_name(fp, target_dir, chain):\n",
    "    orig_path, file_name = fp.rsplit(\"/\", 1)\n",
    "    fp = fp.replace(orig_path, target_dir)\n",
    "    fp = fp.split(\".\")[0] + f\"_{chain}.pdb\"\n",
    "    return fp\n",
    "\n",
    "def find_pdb_file(PDB_chain_id, files):\n",
    "    print(files)\n",
    "    if \"_\" in PDB_chain_id:\n",
    "        PDB_id = PDB_chain_id.split(\"_\")[0]\n",
    "    assert len(PDB_id) == 4\n",
    "    for f in files:\n",
    "        filename = os.path.basename(f).split('.')[0]\n",
    "        if filename.startswith(PDB_id):\n",
    "            print(f\"Found file: {f} for PDB ID: {PDB_id}\")\n",
    "            return f\n",
    "\n",
    "print(\"BEGIN\")\n",
    "t1 = t()\n",
    "xml = \"braf-blast.xml\"\n",
    "pdb_dir = \"PDBs\"\n",
    "target_dir = \"Results/activation_segments/unaligned\"\n",
    "\n",
    "# Ensure the target directory exists\n",
    "os.makedirs(target_dir, exist_ok=True)\n",
    "\n",
    "xml_chain_dict = parse_xml(xml)\n",
    "print(xml_chain_dict)\n",
    "\n",
    "pdb_files = sorted(sglob(\"PDBs/*.pdb\"))\n",
    "print(pdb_files)\n",
    "keys = sorted([*xml_chain_dict.keys()], key=get_pdb_id)\n",
    "\n",
    "print(keys)\n",
    "files = [find_pdb_file(k, pdb_files) for k in keys]\n",
    "print(files)\n",
    "chain_IDs = [xml_chain_dict[k] for k in keys]\n",
    "print(len(chain_IDs))\n",
    "\n",
    "valid_file_chain_pairs = [(f, c) for f, c in zip(files, chain_IDs) if f is not None]\n",
    "print(len(valid_file_chain_pairs))\n",
    "new_file_paths = [target_name(f, target_dir, c) for f, c in valid_file_chain_pairs]\n",
    "print(new_file_paths)\n",
    "file_paths = [f for f, c in valid_file_chain_pairs]\n",
    "print(file_paths)\n",
    "print(len(file_paths))\n",
    "\n",
    "for (fp, chain_ID, tp) in zip(file_paths, chain_IDs, new_file_paths):\n",
    "    if fp is not None:\n",
    "        try:\n",
    "            print(fp, chain_ID, tp)\n",
    "            chain = strip_to_chain(fp, chain_ID)\n",
    "            if chain is not None:\n",
    "                with mda.Writer(tp) as w:\n",
    "                    w.write(chain)\n",
    "                post_process(tp)\n",
    "        except Exception as e:\n",
    "            print(f\"An error occurred while processing {fp} with chain {chain_ID}: {e}\")\n",
    "        \n",
    "t2 = t()\n",
    "t_t = round(t2 - t1, 3) // 60\n",
    "t_t = str((t_t // 60)) + \":\" + str(t_t % 60)\n",
    "print(f\"Time taken {t_t} for sequential processing\")\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "616153e7",
   "metadata": {},
   "source": [
    "## Counting again how many pdb files are in the directory after stripping the chains"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "3b69d1ac",
   "metadata": {},
   "outputs": [],
   "source": [
    "import glob\n",
    "\n",
    "def count_pdb_files(directory):\n",
    "    # Ensure the directory path ends with a slash\n",
    "    directory = os.path.join(directory, '')\n",
    "\n",
    "    # Use glob to find all .pdb files in the directory\n",
    "    pdb_files = glob.glob(os.path.join(directory, '*.pdb'))\n",
    "\n",
    "    # Return the count of .pdb files\n",
    "    return len(pdb_files)\n",
    "\n",
    "# Specify the directory\n",
    "pdb_directory = 'Results/activation_segments/unaligned'\n",
    "\n",
    "# Get the count of PDB files\n",
    "pdb_count = count_pdb_files(pdb_directory)\n",
    "\n",
    "print(f\"There are {pdb_count} PDB files in the directory '{pdb_directory}'.\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "0e014c4b",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Extracting the full sequence from the pdb files, checking if there are any non-natural amino acids and substituting them and selecting only sequences with a maximum gap length of 4 amino acids to be reconstructed\n",
    "from Bio.PDB import PDBParser, PPBuilder\n",
    "from Bio.SeqUtils import seq1\n",
    "from glob import glob\n",
    "from tqdm import tqdm\n",
    "import os\n",
    "\n",
    "def substitute_non_natural_amino_acid(residue, aligned_atom, index):\n",
    "    \"\"\"Substitute non-natural amino acids with their natural counterparts.\"\"\"\n",
    "    substitutions = {\n",
    "        'X': 'G',  # Glycine\n",
    "        'B': 'N',  # Asparagine\n",
    "        'Z': 'Q',  # Glutamine\n",
    "        'J': 'L'   # Leucine\n",
    "    }\n",
    "    \n",
    "    if residue == 'X':\n",
    "        # Check if 'X' is surrounded by missing amino acids in aligned_atom\n",
    "        if index > 0 and aligned_atom[index - 1] == '-':\n",
    "            return residue\n",
    "        if index < len(aligned_atom) - 1 and aligned_atom[index + 1] == '-':\n",
    "            return residue\n",
    "\n",
    "    return substitutions.get(residue, residue)\n",
    "\n",
    "def extract_seqres_sequence(pdb_file):\n",
    "    \"\"\"Extract SEQRES sequences for each chain from a PDB file.\"\"\"\n",
    "    seq_dict = {}\n",
    "    with open(pdb_file, \"r\") as file:\n",
    "        lines = file.readlines()\n",
    "\n",
    "    current_chain = None\n",
    "    current_seq = []\n",
    "\n",
    "    for line in lines:\n",
    "        if line.startswith(\"SEQRES\"):\n",
    "            parts = line.split()\n",
    "            chain_id = parts[2]\n",
    "            if chain_id != current_chain:\n",
    "                if current_chain is not None:\n",
    "                    seq_dict[current_chain] = ''.join(seq1(residue) for residue in current_seq)\n",
    "                current_chain = chain_id\n",
    "                current_seq = []\n",
    "            current_seq.extend(parts[4:])\n",
    "\n",
    "    if current_chain is not None:\n",
    "        seq_dict[current_chain] = ''.join(seq1(residue) for residue in current_seq)\n",
    "\n",
    "    return seq_dict\n",
    "\n",
    "def extract_atom_sequence(pdb_file, chain_id):\n",
    "    \"\"\"Extract sequence from atomic coordinates for a specific chain.\"\"\"\n",
    "    parser = PDBParser(QUIET=True)\n",
    "    structure = parser.get_structure('PDB', pdb_file)\n",
    "    \n",
    "    for model in structure:\n",
    "        chain = model[chain_id]\n",
    "        \n",
    "        ppb = PPBuilder()\n",
    "        sequence = ''\n",
    "        for pp in ppb.build_peptides(chain):\n",
    "            sequence += pp.get_sequence()\n",
    "        return sequence\n",
    "    return None\n",
    "\n",
    "def find_motif_indices(sequence, motif):\n",
    "    \"\"\"Find the start index of a motif in a sequence.\"\"\"\n",
    "    index = sequence.find(motif)\n",
    "    return index if index != -1 else None\n",
    "\n",
    "def align_and_highlight_gaps(seqres_segment, atom_segment):\n",
    "    \"\"\"Align SEQRES and ATOM segments and highlight gaps in ATOM.\"\"\"\n",
    "    aligned_seqres = ''\n",
    "    aligned_atom = ''\n",
    "    atom_index = 0\n",
    "    max_gap_length = 0\n",
    "    current_gap_length = 0\n",
    "\n",
    "    for res_seqres in seqres_segment:\n",
    "        if atom_index < len(atom_segment) and res_seqres == atom_segment[atom_index]:\n",
    "            aligned_seqres += res_seqres\n",
    "            aligned_atom += atom_segment[atom_index]\n",
    "            atom_index += 1\n",
    "            current_gap_length = 0\n",
    "        else:\n",
    "            aligned_seqres += res_seqres\n",
    "            aligned_atom += '-'\n",
    "            current_gap_length += 1\n",
    "            max_gap_length = max(max_gap_length, current_gap_length)\n",
    "\n",
    "    return aligned_seqres, aligned_atom, max_gap_length\n",
    "\n",
    "def main():\n",
    "    target_dir = \"Results/activation_segments/unaligned\"\n",
    "    pdb_dir = \"PDBs\"\n",
    "    fasta_output_file = \"seqres_sequences.fasta\"  # File to store full sequences\n",
    "    text_output_file = \"seqres_info.txt\"\n",
    "    aligned_sequences = {}\n",
    "    satisfying_structures_count = 0\n",
    "\n",
    "    pdb_files = glob(os.path.join(target_dir, \"*.pdb\"))\n",
    "\n",
    "    with open(text_output_file, \"w\") as text_output, open(fasta_output_file, \"w\") as fasta_output:\n",
    "        for pdb_file in tqdm(pdb_files, desc=\"Processing PDB files\"):\n",
    "            pdb_name = os.path.basename(pdb_file)\n",
    "            pdb_id, chain_id_file = os.path.splitext(pdb_name)[0].split('_')\n",
    "            chain_id = chain_id_file\n",
    "\n",
    "            full_pdb_path = os.path.join(pdb_dir, pdb_id + '.pdb')\n",
    "            if not os.path.isfile(full_pdb_path):\n",
    "                print(f\"Corresponding full PDB for {pdb_id} not found.\")\n",
    "                continue\n",
    "\n",
    "            seqres_seqs = extract_seqres_sequence(full_pdb_path)\n",
    "            atom_seq = extract_atom_sequence(full_pdb_path, chain_id)\n",
    "\n",
    "            if chain_id in seqres_seqs and atom_seq:\n",
    "                seqres_sequence = seqres_seqs[chain_id]\n",
    "                seqres_dfg_index = find_motif_indices(seqres_sequence, 'DFG')\n",
    "                seqres_ape_index = find_motif_indices(seqres_sequence, 'APE')\n",
    "                atom_dfg_index = find_motif_indices(atom_seq, 'DFG')\n",
    "                atom_ape_index = find_motif_indices(atom_seq, 'APE')\n",
    "\n",
    "                # Determine the start and end indices for the segments\n",
    "                if None not in [seqres_dfg_index, seqres_ape_index, atom_dfg_index, atom_ape_index]:\n",
    "                    seqres_start = min(seqres_dfg_index, seqres_ape_index)\n",
    "                    seqres_end = max(seqres_dfg_index + 3, seqres_ape_index + 3)\n",
    "                    atom_start = min(atom_dfg_index, atom_ape_index)\n",
    "                    atom_end = max(atom_dfg_index + 3, atom_ape_index + 3)\n",
    "\n",
    "                    seqres_segment = seqres_sequence[seqres_start:seqres_end]\n",
    "                    atom_segment = atom_seq[atom_start:atom_end]\n",
    "\n",
    "                    aligned_seqres, aligned_atom, max_gap_length = align_and_highlight_gaps(seqres_segment, atom_segment)\n",
    "                    \n",
    "                    # Check for differences and substitute non-natural amino acids\n",
    "                    exclude_due_to_non_natural_diff = False\n",
    "                    corrected_seqres = ''\n",
    "                    for index, (res_seqres, res_atom) in enumerate(zip(aligned_seqres, aligned_atom)):\n",
    "                        if res_seqres != res_atom:\n",
    "                            corrected_residue = substitute_non_natural_amino_acid(res_seqres, aligned_atom, index)\n",
    "                            corrected_seqres += corrected_residue\n",
    "                            if corrected_residue != res_seqres:\n",
    "                                print(f\"Substituting non-natural amino acid '{res_seqres}' with '{corrected_residue}' in SEQRES for {pdb_id}_{chain_id}.\")\n",
    "                        else:\n",
    "                            corrected_seqres += res_seqres\n",
    "\n",
    "                    if not exclude_due_to_non_natural_diff and max_gap_length <= 4:\n",
    "                        satisfying_structures_count += 1\n",
    "                        info = (f\"Aligned Sequences for {pdb_id}_{chain_id}: (Max gap length: {max_gap_length})\\n\"\n",
    "                                f\"SEQRES Segment: {corrected_seqres}\\n\"\n",
    "                                f\"ATOM Segment:   {aligned_atom}\\n\\n\")\n",
    "                        print(info)\n",
    "                        text_output.write(info)\n",
    "                        \n",
    "                        aligned_sequences[f\"{pdb_id}_{chain_id}_SEQRES\"] = corrected_seqres\n",
    "                        aligned_sequences[f\"{pdb_id}_{chain_id}_ATOM\"] = aligned_atom\n",
    "\n",
    "                        # Write full SEQRES and ATOM sequences to the FASTA file\n",
    "                        fasta_output.write(f\">{pdb_id}_{chain_id}_SEQRES\\n{seqres_sequence}\\n\")\n",
    "                        fasta_output.write(f\">{pdb_id}_{chain_id}_ATOM\\n{atom_seq}\\n\")\n",
    "                    else:\n",
    "                        exclusion_msg = f\"Excluding {pdb_id}_{chain_id} due to gap length: {max_gap_length} or non-natural amino acid difference.\\n\"\n",
    "                        print(exclusion_msg)\n",
    "                        text_output.write(exclusion_msg)\n",
    "                else:\n",
    "                    motif_msg = f\"Motifs not found in {pdb_id}_{chain_id}.\\n\"\n",
    "                    print(motif_msg)\n",
    "                    text_output.write(motif_msg)\n",
    "            else:\n",
    "                chain_msg = f\"Chain {chain_id} not found in SEQRES of {pdb_id} or no atomic sequence available.\\n\"\n",
    "                print(chain_msg)\n",
    "                text_output.write(chain_msg)\n",
    "                \n",
    "        count_msg = f\"Total structures satisfying the condition: {satisfying_structures_count}\"\n",
    "        text_output.write(count_msg)\n",
    "        print(count_msg)\n",
    "\n",
    "if __name__ == '__main__':\n",
    "    main()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "f3c5b45f",
   "metadata": {},
   "outputs": [],
   "source": [
    "#COunting how many seqences are in the fasta file\n",
    "def count_total_pdb_ids(file_path):\n",
    "    total_pdb_ids = 0\n",
    "\n",
    "    with open(file_path, 'r') as file:\n",
    "        for line in file:\n",
    "            if line.startswith('>'):\n",
    "                total_pdb_ids += 1\n",
    "\n",
    "    print(f\"Total number of PDB IDs: {int(total_pdb_ids/2)}\") #here we divide by 2 because we have two lines per PDB ID\n",
    "\n",
    "# Provide the path to your seqres_sequence.fasta file\n",
    "file_path = \"seqres_sequences.fasta\"\n",
    "count_total_pdb_ids(file_path)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "844a44c7",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Code to use MODELLER to reconstruct the sequences that have a gap length of 4 or less, if there are no differences between the SEQRES and ATOM sequences, the original PDB file is copied to the target directory\n",
    "# FIX was to eliminate remarks from the pdb file\n",
    "import os\n",
    "import shutil\n",
    "from Bio.PDB import PDBParser, PPBuilder\n",
    "from modeller import *\n",
    "from modeller.automodel import *\n",
    "\n",
    "def read_fasta_sequences(fasta_file):\n",
    "    \"\"\"Read sequences from a FASTA file into a dictionary.\"\"\"\n",
    "    sequences = {}\n",
    "    with open(fasta_file, \"r\") as f:\n",
    "        lines = f.readlines()\n",
    "        header = None\n",
    "        sequence = []\n",
    "        for line in lines:\n",
    "            line = line.strip()\n",
    "            if line.startswith(\">\"):\n",
    "                if header:\n",
    "                    sequences[header] = ''.join(sequence)\n",
    "                header = line[1:]\n",
    "                sequence = []\n",
    "            else:\n",
    "                sequence.append(line)\n",
    "        if header:\n",
    "            sequences[header] = ''.join(sequence)\n",
    "    return sequences\n",
    "\n",
    "def extract_atom_sequence(pdb_file):\n",
    "    \"\"\"Extract sequence from atomic coordinates for the first chain found in the PDB file.\"\"\"\n",
    "    parser = PDBParser(QUIET=True)\n",
    "    structure = parser.get_structure('PDB', pdb_file)\n",
    "    \n",
    "    for model in structure:\n",
    "        for chain in model:\n",
    "            ppb = PPBuilder()\n",
    "            sequence = ''\n",
    "            for pp in ppb.build_peptides(chain):\n",
    "                sequence += pp.get_sequence()\n",
    "            return str(sequence)\n",
    "    return None\n",
    "\n",
    "def find_motif_indices(sequence, motif):\n",
    "    \"\"\"Find the start index of a motif in a sequence.\"\"\"\n",
    "    index = sequence.find(motif)\n",
    "    return index if index != -1 else None\n",
    "\n",
    "def reconstruct_with_modeller(pdb_chain_id, pdb_path, target_path, full_sequence, atom_sequence):\n",
    "    print(f\"Processing {pdb_chain_id}\")\n",
    "\n",
    "    # Find indices of the DFG and APE motifs\n",
    "    seqres_dfg_index = find_motif_indices(full_sequence, 'DFG')\n",
    "    seqres_ape_index = find_motif_indices(full_sequence, 'APE')\n",
    "    atom_dfg_index = find_motif_indices(atom_sequence, 'DFG')\n",
    "    atom_ape_index = find_motif_indices(atom_sequence, 'APE')\n",
    "\n",
    "    # Determine if reconstruction is needed\n",
    "    if None not in [seqres_dfg_index, seqres_ape_index, atom_dfg_index, atom_ape_index]:\n",
    "        seqres_start = min(seqres_dfg_index, seqres_ape_index)\n",
    "        seqres_end = max(seqres_dfg_index + 3, seqres_ape_index + 3)\n",
    "        atom_start = min(atom_dfg_index, atom_ape_index)\n",
    "        atom_end = max(atom_dfg_index + 3, atom_ape_index + 3)\n",
    "\n",
    "        seqres_segment = full_sequence[seqres_start:seqres_end]\n",
    "        atom_segment = atom_sequence[atom_start:atom_end]\n",
    "\n",
    "        # Check for differences in the segment\n",
    "        if seqres_segment != atom_segment:\n",
    "            print(f\"Reconstructing full sequence for {pdb_chain_id} using MODELLER\")\n",
    "\n",
    "            # Setting up MODELLER\n",
    "            env = environ()\n",
    "            aln = alignment(env)\n",
    "            \n",
    "            # Read the structure to work on\n",
    "            mdl = model(env, file=pdb_path)\n",
    "            aln.append_model(mdl, align_codes='template', atom_files=pdb_path)\n",
    "\n",
    "            # Append the full target sequence\n",
    "            aln.append_sequence(full_sequence)\n",
    "            aln[-1].code = 'target'\n",
    "            \n",
    "            # Perform the alignment\n",
    "            aln.align2d(max_gap_length=50)\n",
    "\n",
    "            # Create AutoModel object and build models\n",
    "            a = automodel(env, alnfile=aln, knowns='template', sequence='target')\n",
    "            a.starting_model = 1\n",
    "            a.ending_model = 1\n",
    "            \n",
    "            # Build the model\n",
    "            a.make()\n",
    "            \n",
    "            # Save the best model to the target directory\n",
    "            model_path = os.path.join(target_path, f\"{pdb_chain_id}_filled.pdb\")\n",
    "            os.rename(a.outputs[0]['name'], model_path)\n",
    "            print(f\"Reconstruction completed for {pdb_chain_id}. File saved at {model_path}\")\n",
    "\n",
    "            # Post-process to remove REMARK lines\n",
    "            remove_remark_lines(model_path)\n",
    "        else:\n",
    "            # No reconstruction needed, copy original PDB\n",
    "            shutil.copy(pdb_path, os.path.join(target_path, f\"{pdb_chain_id}.pdb\"))\n",
    "            print(f\"No differences found for {pdb_chain_id}. Original PDB copied to target directory.\")\n",
    "    else:\n",
    "        print(f\"Motifs not found in {pdb_chain_id}.\")\n",
    "\n",
    "def remove_remark_lines(pdb_file):\n",
    "    \"\"\"Remove lines starting with 'REMARK' from the PDB file.\"\"\"\n",
    "    with open(pdb_file, 'r') as file:\n",
    "        lines = file.readlines()\n",
    "    \n",
    "    with open(pdb_file, 'w') as file:\n",
    "        for line in lines:\n",
    "            if not line.startswith(\"REMARK\"):\n",
    "                file.write(line)\n",
    "    \n",
    "    print(f\"Removed REMARK lines from {pdb_file}\")\n",
    "\n",
    "def main():\n",
    "    seqres_fasta = \"seqres_sequences.fasta\"\n",
    "    pdb_dir = \"Results/activation_segments/unaligned\"\n",
    "    target_dir = \"Results/activation_segments/reconstructedModeller\"\n",
    "\n",
    "    # Read the sequences from the FASTA file\n",
    "    seqres_sequences = read_fasta_sequences(seqres_fasta)\n",
    "\n",
    "    for header, full_sequence in seqres_sequences.items():\n",
    "        if \"_SEQRES\" in header:  # Only consider SEQRES entries\n",
    "            pdb_chain_id = header.replace(\"_SEQRES\", \"\")\n",
    "            pdb_file_path = os.path.join(pdb_dir, f\"{pdb_chain_id}.pdb\")\n",
    "\n",
    "            # Extract the atomic sequence\n",
    "            atom_sequence = extract_atom_sequence(pdb_file_path)\n",
    "\n",
    "            if atom_sequence is None:\n",
    "                print(f\"Could not extract sequence for {pdb_chain_id}. Skipping...\")\n",
    "                continue\n",
    "\n",
    "            if not os.path.exists(target_dir):\n",
    "                os.makedirs(target_dir)\n",
    "            \n",
    "            reconstruct_with_modeller(pdb_chain_id, pdb_file_path, target_dir, full_sequence, atom_sequence)\n",
    "\n",
    "    print(\"Processing complete!\")\n",
    "\n",
    "if __name__ == '__main__':\n",
    "    main()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "fccfa004",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Counting the number of pdb files in the directory after reconstruction\n",
    "import glob\n",
    "import os\n",
    "def count_pdb_files(directory):\n",
    "    # Ensure the directory path ends with a slash\n",
    "    directory = os.path.join(directory, '')\n",
    "\n",
    "    # Use glob to find all .pdb files in the directory\n",
    "    pdb_files = glob.glob(os.path.join(directory, '*.pdb'))\n",
    "\n",
    "    # Return the count of .pdb files\n",
    "    return len(pdb_files)\n",
    "\n",
    "# Specify the directory\n",
    "pdb_directory = 'Results/activation_segments/reconstructedModeller'\n",
    "\n",
    "# Get the count of PDB files\n",
    "pdb_count = count_pdb_files(pdb_directory)\n",
    "\n",
    "print(f\"There are {pdb_count} PDB files in the directory '{pdb_directory}'.\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "119237e7",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Fast checking if reconstruction was successful, need to just change the pdb id and chain id\n",
    "import MDAnalysis as mda\n",
    "import nglview as nv\n",
    "from Bio.PDB import PDBParser\n",
    "from MDAnalysis.analysis import align\n",
    "\n",
    "# Dictionary to convert three-letter amino acid codes to one-letter codes\n",
    "three_to_one = {\n",
    "    'ALA': 'A', 'ARG': 'R', 'ASN': 'N', 'ASP': 'D',\n",
    "    'CYS': 'C', 'GLU': 'E', 'GLN': 'Q', 'GLY': 'G',\n",
    "    'HIS': 'H', 'ILE': 'I', 'LEU': 'L', 'LYS': 'K',\n",
    "    'MET': 'M', 'PHE': 'F', 'PRO': 'P', 'SER': 'S',\n",
    "    'THR': 'T', 'TRP': 'W', 'TYR': 'Y', 'VAL': 'V'\n",
    "}\n",
    "\n",
    "def extract_sequence_and_mapping(pdb_file):\n",
    "    \"\"\"Extract sequence and create a mapping from sequence index to PDB residue ID.\"\"\"\n",
    "    parser = PDBParser(QUIET=True)\n",
    "    structure = parser.get_structure('PDB', pdb_file)\n",
    "    \n",
    "    sequence = []\n",
    "    index_to_resid = {}\n",
    "    \n",
    "    for model in structure:\n",
    "        for chain in model:\n",
    "            for residue in chain:\n",
    "                if 'CA' in residue:  # Check if it's an amino acid\n",
    "                    resname = residue.get_resname()\n",
    "                    if resname in three_to_one:\n",
    "                        sequence.append(three_to_one[resname])\n",
    "                        index_to_resid[len(sequence) - 1] = residue.get_id()[1]  # Map sequence index to PDB resid\n",
    "            break\n",
    "        break\n",
    "\n",
    "    return sequence, index_to_resid\n",
    "\n",
    "def find_motif_indices(sequence, motif):\n",
    "    \"\"\"Find the start index of a motif in a sequence.\"\"\"\n",
    "    sequence_str = ''.join(sequence)\n",
    "    index = sequence_str.find(motif)\n",
    "    return index if index != -1 else None\n",
    "\n",
    "# Extract the sequence and mapping for the single chain\n",
    "atom_sequence, index_to_resid = extract_sequence_and_mapping(\"Results/activation_segments/unaligned/7OPO_A.pdb\")\n",
    "\n",
    "# Find indices of DFG and APE motifs\n",
    "dfg_index = find_motif_indices(atom_sequence, 'DFG')\n",
    "ape_index = find_motif_indices(atom_sequence, 'APE')\n",
    "\n",
    "# Ensure indices are found and select the residues between them\n",
    "if dfg_index is not None and ape_index is not None:\n",
    "    # Use the mapping to get the correct residue IDs\n",
    "    dfg_resid = index_to_resid[dfg_index]\n",
    "    ape_resid = index_to_resid[ape_index + 2]  # +2 to include the entire 'APE' motif\n",
    "\n",
    "    u_missing = mda.Universe(\"Results/activation_segments/unaligned/7OPO_A.pdb\")\n",
    "    selected_atoms = u_missing.select_atoms(f\"resid {dfg_resid}:{ape_resid}\")\n",
    "\n",
    "    print(\"Number of Atoms Selected:\", selected_atoms.n_atoms)\n",
    "\n",
    "    u_reconstructed = mda.Universe(\"Results/activation_segments/reconstructedModeller/7OPO_A_filled.pdb\")\n",
    "    print(\"Number of Atoms Reconstructed:\", u_reconstructed.select_atoms(f\"all\").n_atoms)\n",
    "\n",
    "    # Merge the aligned atoms for visualization\n",
    "    merged = mda.Merge(selected_atoms, u_reconstructed.atoms)\n",
    "    print(merged.residues)\n",
    "\n",
    "    # Create NGLView widget\n",
    "    w = nv.show_mdanalysis(merged)\n",
    "\n",
    "    # Add a representation for each residue name with the corresponding color\n",
    "    w.clear()\n",
    "    w.add_cartoon(color=\"resname\")\n",
    "\n",
    "    \n",
    "\n",
    "else:\n",
    "    print(\"Motifs not found in the sequence.\")\n",
    "\n",
    "w"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "766c8fc9",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Helper functions and function to run MUSTANG on the reconstructed pdb files\n",
    "\n",
    "import subprocess\n",
    "import os\n",
    "from glob import glob as g\n",
    "import mdtraj as md\n",
    "from mpi4py import MPI\n",
    "from time import time as t\n",
    "from tqdm import tqdm\n",
    "\n",
    "\n",
    "def sg(f_p):\n",
    "    return sorted(g(f_p))\n",
    "\n",
    "\n",
    "def find_pdbs(directory):\n",
    "    \"\"\"\n",
    "    Find topologies in a directory.\n",
    "    Currently excludes cif files.\n",
    "    \"\"\"\n",
    "    return sg(directory+\"/*.pdb\")\n",
    "\n",
    "\n",
    "def fname(file):\n",
    "    return file.rsplit(\".\", 1)[0].rsplit(\"/\", 1)[-1]\n",
    "\n",
    "\n",
    "def ifnotmake(dir_path):\n",
    "    if not os.path.isdir(dir_path):\n",
    "        os.makedirs(dir_path)\n",
    "    return dir_path\n",
    "\n",
    "\n",
    "def run_mustang(f1, f2, name=None):\n",
    "    \"\"\"\n",
    "    Writes a MUSTANG input file which aligns\n",
    "    file1 to file 2.\n",
    "    If no name defaults to the second file.\n",
    "    \"\"\"\n",
    "    if name is None:\n",
    "        name = fname(f2)\n",
    "    if not os.path.isdir(f\"./{name}\"):\n",
    "        os.makedirs(f\"./{name}\")\n",
    "    new_fp = f\"./{name}/{name}\"\n",
    "    structs = f\"{f1} {f2} \"\n",
    "    command = f\"/home/marmatt/Downloads/MUSTANG_v3.2.4/bin/mustang-3.2.4 -i {structs} -o {new_fp} -F fasta -s ON\" #here you should change the path to where you install MUSTANG\n",
    "    command = command.split()\n",
    "    new_fp = f\"{new_fp}.pdb\"\n",
    "    try:\n",
    "        result = subprocess.run(command, capture_output=True, text=True)\n",
    "        print(f\"Running command: {' '.join(command)}\")\n",
    "        print(f\"STDOUT: {result.stdout}\")\n",
    "        print(f\"STDERR: {result.stderr}\")\n",
    "        if result.returncode != 0:\n",
    "            print(\"Error in MUSTANG execution\")\n",
    "            return None\n",
    "    except Exception as e:\n",
    "        print(f\"Exception occurred: {e}\")\n",
    "        return None\n",
    "    return new_fp\n",
    "\n",
    "\n",
    "def postprocess(file_path):\n",
    "    \"\"\"\n",
    "    file_path is the name of a pdb file.\n",
    "    It deletes the first chain which is always the alignment structures\n",
    "    \"\"\"\n",
    "    structure = md.load(file_path)\n",
    "    aligned_chain_idx = [[atom.index for atom in res.atoms] for res in\n",
    "                         structure.top._chains[1]._residues]\n",
    "    aligned_chain_idx = sum(aligned_chain_idx, [])\n",
    "    structure = structure.atom_slice(aligned_chain_idx)\n",
    "    structure.save(file_path)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "993598b0",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Perform MUSTANG alignment on the sequences that have not been reconstructed, I need to fix this bug\n",
    "\n",
    "#from mustang import *\n",
    "import subprocess\n",
    "from mpi4py import MPI\n",
    "\n",
    "# Constants (most to be made variable)\n",
    "\n",
    "comm = MPI.COMM_WORLD\n",
    "rank = comm.Get_rank()\n",
    "size = comm.Get_size()\n",
    "# Define the paths explicitly\n",
    "pdb_path = \"Results/activation_segments/reconstructedModeller\"\n",
    "#pdb_path = \"/home/marmatt/Documents/projects/BRAF/myWork/reproduceBRAFWork/Results/activation_segments/unaligned\"\n",
    "target_dir = \"Results/activation_segments/mustangs\"\n",
    "template_pdb = \"6UAN_chainD.pdb\"\n",
    "# Ensure the target directory exists\n",
    "ifnotmake(target_dir)\n",
    "pdb_path = os.path.abspath(pdb_path)\n",
    "target_dir = os.path.abspath(target_dir)\n",
    "template_pdb = os.path.abspath(template_pdb)\n",
    "print(pdb_path, target_dir, template_pdb)\n",
    "os.chdir(target_dir)\n",
    "os.system(\"pwd\")\n",
    "if rank == 0:\n",
    "    pdbs = find_pdbs(pdb_path)\n",
    "    n_files = len(pdbs)\n",
    "    n_slices = (n_files // size)\n",
    "    step = int(n_files / n_slices)\n",
    "    if n_files % n_slices != 0:\n",
    "        n_slices += 1\n",
    "    slices = [slice(i*n_slices, (i+1)*n_slices) for i in range(step)]\n",
    "    pdbs = [pdbs[s] for s in slices]\n",
    "else:\n",
    "    pdbs = None\n",
    "\n",
    "pdbs = comm.scatter(pdbs, root=0)\n",
    "print(\"RANK:\\t\", rank, \"DATA SIZE:\\t\", len(pdbs))\n",
    "t1 = t()\n",
    "failures = []\n",
    "for pdb in tqdm(pdbs):\n",
    "    name = fname(pdb)\n",
    "    new_fp = run_mustang(template_pdb, pdb, name=name)\n",
    "    if new_fp:\n",
    "        if os.path.isfile(new_fp):\n",
    "            postprocess(new_fp)\n",
    "        else:\n",
    "            failures.append(pdb)\n",
    "    else:\n",
    "        failures.append(pdb)\n",
    "t2 = t()\n",
    "print(\"FINISHED RANK:\\t\", rank, \"DATA SIZE:\\t\", len(pdbs),\n",
    "      \"TIME:\\t\", round(t2-t1, 4))\n",
    "failures = comm.gather(failures, root=0)\n",
    "if rank == 0:\n",
    "    failures = sum(failures, [])\n",
    "    with open(\"./failures.txt\", \"w\") as f_o:\n",
    "        f_o.write(\"\\n\".join(f for f in failures))\n",
    "    t2 = t()\n",
    "    print(round(t2-t1, 3))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "577969e9",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Code to truncate the aligned structures to only activation segment and save them to pdb files\n",
    "'''\n",
    "import os\n",
    "import mdtraj as md\n",
    "from glob import glob as g\n",
    "\n",
    "def ifnotmake(dir_path):\n",
    "    if not os.path.isdir(dir_path):\n",
    "        os.makedirs(dir_path)\n",
    "    return dir_path\n",
    "\n",
    "def sg(f_p):\n",
    "    return sorted(g(f_p))\n",
    "\n",
    "def find_pdbs(directory):\n",
    "    \"\"\"\n",
    "    Find PDB files in a directory and its subdirectories.\n",
    "    \"\"\"\n",
    "    return sg(directory + \"/**/*.pdb\")\n",
    "\n",
    "def find_motif_indices(sequence, motif):\n",
    "    \"\"\"\n",
    "    Find the starting index of a motif in a sequence.\n",
    "    Returns the index of the first residue of the motif.\n",
    "    \"\"\"\n",
    "    motif_length = len(motif)\n",
    "    for i in range(len(sequence) - motif_length + 1):\n",
    "        if sequence[i:i + motif_length] == motif:\n",
    "            return i\n",
    "    return None\n",
    "\n",
    "def truncate_pdb(file_path, output_dir, motif_start='DFG', motif_end='APE'):\n",
    "    \"\"\"\n",
    "    Truncate the PDB file between the specified motifs and save to output directory.\n",
    "    \"\"\"\n",
    "    # Load the structure\n",
    "    structure = md.load(file_path)\n",
    "    sequence = [res.name for res in structure.topology.residues]\n",
    "\n",
    "    # Find indices of the motifs\n",
    "    start_idx = find_motif_indices(sequence, motif_start)\n",
    "    end_idx = find_motif_indices(sequence, motif_end)\n",
    "\n",
    "    if start_idx is None or end_idx is None:\n",
    "        print(f\"Motifs not found in {file_path}\")\n",
    "        return\n",
    "\n",
    "    # Ensure the end motif is after the start motif\n",
    "    if start_idx >= end_idx:\n",
    "        print(f\"Invalid motif order in {file_path}\")\n",
    "        return\n",
    "\n",
    "    # Slice the structure\n",
    "    truncated_structure = structure.atom_slice(structure.top.select(f\"residue >= {start_idx} and residue <= {end_idx}\"))\n",
    "\n",
    "    # Extract PDB ID from the file name\n",
    "    pdb_id = os.path.basename(file_path).split('.')[0]\n",
    "\n",
    "    # Save the truncated PDB with the PDB ID\n",
    "    output_path = os.path.join(output_dir, f\"{pdb_id}_truncated.pdb\")\n",
    "    truncated_structure.save(output_path)\n",
    "    print(f\"Truncated PDB saved to {output_path}\")\n",
    "\n",
    "def process_pdbs(input_dir, output_dir):\n",
    "    \"\"\"\n",
    "    Process all PDB files in the input directory and save truncated versions to the output directory.\n",
    "    \"\"\"\n",
    "    ifnotmake(output_dir)\n",
    "    pdb_files = find_pdbs(input_dir)\n",
    "\n",
    "    for pdb_file in pdb_files:\n",
    "        truncate_pdb(pdb_file, output_dir)\n",
    "\n",
    "# Define paths\n",
    "input_directory = \"Results/activation_segments/mustangs\"\n",
    "output_directory = \"Results/activation_segments/mustang_truncated\"\n",
    "\n",
    "# Ensure output directory exists\n",
    "ifnotmake(output_directory)\n",
    "\n",
    "# Process all PDB files\n",
    "process_pdbs(input_directory, output_directory)\n",
    "'''"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "7edb73e8",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Code to save only first and last residue in activation segment for further visualization\n",
    "\n",
    "import biobox as bb\n",
    "import glob\n",
    "import numpy as np\n",
    "\n",
    "files = \"Results/activation_segments/mustang_truncated/*.pdb\"\n",
    "\n",
    "start = []\n",
    "end []\n",
    "for f in glob.glob(files):\n",
    "\n",
    "    M = bb.Molecule(f)\n",
    "    pts = M.atomselect(\"*\", \"*\", \"CA\")\n",
    "\n",
    "    start.append(pts[0])\n",
    "    end.append(pts[-1])\n",
    "\n",
    "start = np.array(start)\n",
    "end = np.array(end)\n",
    "\n",
    "M_start = bb.Structure(pts=start)\n",
    "M_end = bb.Structure(pts=end)\n",
    "\n",
    "M_start.write_pdb(\"start_CA.pdb\")\n",
    "M_end.write_pdb(\"end_CA.pdb\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "14388f28",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Counting the number of directories representing the number of pdb files that have been aligned\n",
    "import os\n",
    "def count_directories(directory):\n",
    "    # List all entries in the given directory\n",
    "    entries = os.listdir(directory)\n",
    "\n",
    "    # Use os.path.join to get the full path and os.path.isdir to check if it's a directory\n",
    "    directories = [entry for entry in entries if os.path.isdir(os.path.join(directory, entry))]\n",
    "\n",
    "    # Return the count of directories\n",
    "    return len(directories)\n",
    "\n",
    "# Specify the directory\n",
    "directory_path = 'Results/activation_segments/mustangs'\n",
    "\n",
    "# Get the count of directories\n",
    "directory_count = count_directories(directory_path)\n",
    "\n",
    "print(f\"There are {directory_count} directories in the directory '{directory_path}'.\")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "a3ce23ad",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Counting the number of files representing the number of pdb files that did not need reconstruction\n",
    "import os\n",
    "from glob import glob as g\n",
    "\n",
    "def count_non_filled_pdbs(directory):\n",
    "    # Find all PDB files in the directory\n",
    "    pdb_files = g(os.path.join(directory, \"*.pdb\"))\n",
    "\n",
    "    # Filter out files with '_filled' in their names\n",
    "    non_filled_pdbs = [pdb for pdb in pdb_files if \"_filled\" not in os.path.basename(pdb)]\n",
    "\n",
    "    # Return the count of non '_filled' PDB files\n",
    "    return len(non_filled_pdbs)\n",
    "\n",
    "# Specify the directory\n",
    "pdb_directory_path = 'Results/activation_segments/reconstructedModeller'\n",
    "\n",
    "# Get the count of non '_filled' PDB files\n",
    "non_filled_pdb_count = count_non_filled_pdbs(pdb_directory_path)\n",
    "\n",
    "print(f\"There are {non_filled_pdb_count} PDB files without '_filled' in the directory '{pdb_directory_path}'.\")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "c47016e2",
   "metadata": {},
   "outputs": [],
   "source": [
    "import re\n",
    "import os\n",
    "import h5py\n",
    "import numpy as np\n",
    "import mdtraj as md\n",
    "import pickle as p\n",
    "from tqdm import tqdm\n",
    "from glob import glob\n",
    "from pprint import pprint as pp\n",
    "alignment_dir = \"/home/marmatt/Documents/projects/BRAF/myWork/reproduceBRAFWork/Results/activation_segments/mustangs\"\n",
    "\n",
    "class alignment:\n",
    "    \"\"\"\n",
    "    Class to hold alignments.\n",
    "    Currently only supports braf_monomers!\n",
    "    \"\"\"\n",
    "    def __init__(self,name,seq1,seq2):\n",
    "        self.name = name\n",
    "        self.seq1 = seq1\n",
    "        self.seq2 = seq2\n",
    "        self.aligned = self.find_aligned()\n",
    "\n",
    "    def find_pdb(self):\n",
    "        pdb_files = []\n",
    "        for root, dirs, files in os.walk(alignment_dir):\n",
    "            pdb_files += [os.path.join(root, file) for file in files if file.endswith('.pdb')]\n",
    "        \n",
    "        pdb = [f for f in pdb_files if self.name in f]\n",
    "        print(pdb)\n",
    "        if len(pdb) == 1:\n",
    "            return pdb[0]\n",
    "        \n",
    "\n",
    "    def find_aligned(self):\n",
    "        aligned = []\n",
    "        for char1, char2 in zip(self.seq1,self.seq2):\n",
    "            if char1 != \"-\":\n",
    "                aligned.append((char1,char2))\n",
    "        return aligned\n",
    "\n",
    "    def aligned_res(self):\n",
    "        seq1, seq2 = self.seq1, self.seq2\n",
    "        aligned = [[*item] for item in self.aligned]\n",
    "        seq_length = len(aligned)\n",
    "        pdb2_top = self.load_pdb()\n",
    "        full_seq2 = \"\".join(char for char in seq2 if char != \"-\")\n",
    "        residues = pdb2_top.top._residues\n",
    "        n_res = len(residues)\n",
    "        res_counter = 0\n",
    "        for i in range(seq_length):\n",
    "            if res_counter >= n_res:\n",
    "                break\n",
    "            if aligned[i][1] != \"-\":\n",
    "                aligned[i][1] = residues[res_counter]\n",
    "                res_counter += 1\n",
    "            else:\n",
    "                continue\n",
    "        return [tuple(a) for a in aligned]\n",
    "\n",
    "    def aligned_xyz(self):\n",
    "        \"\"\"\n",
    "        Return xyz of aligned residues\n",
    "        \"\"\"\n",
    "        xyz = self.load_pdb()._xyz[0] # Only one frame\n",
    "        aligned = [[*item] for item in self.aligned]\n",
    "        for k,(_,res) in enumerate(self.residues):\n",
    "            if not isinstance(res,str):\n",
    "                idxs = []\n",
    "                for atom in res._atoms:\n",
    "                    idxs.append(atom.index)\n",
    "                res_xyz = xyz[idxs]\n",
    "                aligned[k][1] = res_xyz\n",
    "            else:\n",
    "                continue\n",
    "        return [tuple(a) for a in aligned]\n",
    "\n",
    "    def aligned_ca_xyz(self):\n",
    "        \"\"\"\n",
    "        Return xyz of aligned residues\n",
    "        \"\"\"\n",
    "        xyz = self.load_pdb()._xyz[0] # Only one frame\n",
    "        aligned = [[*item] for item in self.aligned]\n",
    "        for k,(_,res) in enumerate(self.residues):\n",
    "            if not isinstance(res,str):\n",
    "                for atom in res._atoms:\n",
    "                    if atom.name == \"CA\":\n",
    "                        idxs = atom.index\n",
    "                        break\n",
    "                try:\n",
    "                    res_xyz = xyz[idxs]\n",
    "                except Exception as e:\n",
    "                    print(e)\n",
    "                    print(\"ERROR FOR:\")\n",
    "                    print(self.name)\n",
    "                    return None\n",
    "                aligned[k][1] = res_xyz\n",
    "            else:\n",
    "                continue\n",
    "        return [tuple(a) for a in aligned]\n",
    "\n",
    "    def load_pdb(self):\n",
    "        return md.load(self.pdb_file)\n",
    "\n",
    "    def __getitem__(self,idx):\n",
    "        return (self.seq1[idx],self.seq2[idx])\n",
    "\n",
    "    def __repr__(self):\n",
    "        return self.name\n",
    "\n",
    "    def find_match_id(self):\n",
    "        seq1, seq2 = self.seq1, self.seq2\n",
    "        full_seq2 = \"\".join(char for char in seq2 if char != \"-\")\n",
    "        pp(full_seq2)\n",
    "        aligned = self.aligned\n",
    "        actv_low = 155\n",
    "        actv_hgh = 181\n",
    "        match_residues = aligned[actv_low:actv_hgh] # These are what we need\n",
    "        seq2_Seq = [a[1] for a in match_residues if a[1] != \"-\"]\n",
    "        begin_idx = 0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "19861f06",
   "metadata": {},
   "outputs": [],
   "source": [
    "def afasta_parse(file):\n",
    "    \"\"\"\n",
    "    Parse mustang afasta format output file.\n",
    "    Returns two lists of equal length\n",
    "    \"\"\"\n",
    "    with open(file,\"r\") as f:\n",
    "        lines = f.readlines()\n",
    "    names = [l.split(\".\")[0][1:] for l in lines if l[0] == \">\"]\n",
    "    for i in range(1,len(lines)):\n",
    "        if lines[i].isspace():\n",
    "            lines[i] = \"BREAK\"\n",
    "            break\n",
    "        elif lines[i][0] == \">\":\n",
    "            lines[i] = \"BREAK\" + lines[i]\n",
    "            break\n",
    "    lines = [l.strip() for l in lines if l[0] != \">\"]\n",
    "    lines = \"\".join(lines)\n",
    "    fastas = lines.split(\"BREAK\")\n",
    "    fastas = [*filter(None,fastas)]\n",
    "    return fastas[0], fastas[1]\n",
    "\n",
    "def load_alignments(kind=\"mustang\"):\n",
    "    if kind==\"mustang\":\n",
    "        ppath = \"/home/marmatt/Documents/projects/BRAF/myWork/reproduceBRAFWork/Results/activation_segments/mustangs/mustang_alignments.fasta\"\n",
    "        #print(\"Loading pickled alignments...\")\n",
    "    elif kind==\"blast\":\n",
    "        ppath = \"blast_alignments.fasta\"\n",
    "    if os.path.isfile(ppath):\n",
    "        with open(ppath,\"rb\") as pickled:\n",
    "            #print(\"Loading pickled alignments...\")\n",
    "            return p.load(pickled)\n",
    "    else:\n",
    "        make_align_pickle()\n",
    "        #print(\"No pickled alignments found. Creating...\")\n",
    "        return load_alignments()\n",
    "    \n",
    "def make_align_pickle(kind=\"mustang\"):\n",
    "    if kind == \"mustang\":\n",
    "\n",
    "        alignments = []\n",
    "\n",
    "        # Iterate over directories in the alignment directory\n",
    "        for directory_name in os.listdir(alignment_dir):\n",
    "            directory_path = os.path.join(alignment_dir, directory_name)\n",
    "\n",
    "            # Ensure we are working with directories\n",
    "            if os.path.isdir(directory_path):\n",
    "                #print(f\"Processing directory: {directory_name}\")\n",
    "                fasta_files = tqdm(glob(os.path.join(directory_path, \"*.afasta\")), desc=f\"Processing {directory_name} .afasta files\")\n",
    "                \n",
    "                for fasta_file in fasta_files:\n",
    "                    name = os.path.splitext(os.path.basename(fasta_file))[0]\n",
    "                    fasta_files.set_description(f\"Working on {name}\")\n",
    "\n",
    "                    # Simulate the alignment logic\n",
    "                    aligned = alignment(name, *afasta_parse(fasta_file))  # Assuming `alignment` and `afasta_parse` are predefined\n",
    "                    alignments.append(aligned)\n",
    "\n",
    "        # Define a path for the output pickle file\n",
    "        ppath = os.path.join(alignment_dir, \"mustang_alignments.fasta\")\n",
    "        with open(ppath, \"wb\") as pickled:\n",
    "            p.dump(alignments, pickled)\n",
    "    \n",
    "    elif kind == \"blast\":\n",
    "        b = BLAST_results()\n",
    "        alignments = []\n",
    "        for k, dicti in tqdm(b.alignments.items(),total=len(b.alignments)):\n",
    "            seq1 = dicti[\"Query\"]\n",
    "            seq2 = dicti[\"Subject\"]\n",
    "            alignments.append(alignment(k,seq1,seq2))\n",
    "        ppath = \"blast_alignments.fasta\"\n",
    "        with open(ppath, \"wb\") as pickled:\n",
    "            p.dump(alignments, pickled)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "40a51087",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "3d5581e3c2c1486c975120e3e1971672",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Processing 1FOT_A_filled .afasta files:   0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "2439a52a89c94580845bb0a01e6400e4",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Processing 1BMK_A .afasta files:   0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "34c1b7bf89fc4bf5a1c4647a505d7b99",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Processing 1OGU_A_filled .afasta files:   0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "2d158f77c6b9406b88ae09ecdec6618d",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Processing 1OPK_A .afasta files:   0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "48cd0cd07afc4df280849fcf39ac4a1a",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Processing 1DI9_A .afasta files:   0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "2feb637a9e49468986588895ca9f0a3d",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Processing 1IA8_A .afasta files:   0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "16fc9ab91d31490098694c7d8099e832",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Processing 1GJO_A .afasta files:   0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "7fbe4a6e77374888a9b52193c5b68ab5",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Processing 1B38_A .afasta files:   0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "9a3ab8a933334e9d9eb07dad96f4b2fd",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Processing 1GY3_A_filled .afasta files:   0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "3fbe982a38504f5d8bad40b6ba57cf9d",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Processing 1CTP_E_filled .afasta files:   0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "c53cf459f55c49ffaf81dfbc76be7e4f",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Processing 1AD5_A .afasta files:   0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "7a13217b12364ce182322c2621878498",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Processing 1APM_E_filled .afasta files:   0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "75b751df16e9444da1ef828ecba66e18",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Processing 1J3H_A_filled .afasta files:   0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "3e01094bbf924bcebd4629178ae3337c",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Processing 1GOL_A .afasta files:   0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "0ec40374b9c643e2990d800dc2ad9b68",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Processing 1OPL_A .afasta files: 0it [00:00, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "bce2ce129f4e422cadfdf11a692815d5",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Processing 1CM8_A_filled .afasta files:   0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "81bb696746f149da8b16be2ad9db97c1",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Processing 1FQ1_B_filled .afasta files:   0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "28fde79139c642fda7164d7bdc181784",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Processing 1E9H_A_filled .afasta files:   0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "13c7033b07314a69bbe473de569972a5",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Processing 1BI7_A .afasta files:   0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "a1fa7b666c8e4577abc717a5ca70d505",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Processing 1OIR_A .afasta files:   0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "52091826a5b04d38ba410283fedc2a1f",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Processing 1M7Q_A .afasta files:   0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "1a40eb5252a84da6aeb05dd6eafc78d3",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Processing 1ATP_E_filled .afasta files:   0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "885e03d9a4bd4926a9a07ab46f907bcf",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Processing 1AGW_A .afasta files:   0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "2614c3815eb240caa8315a889c808971",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Processing 1O6K_A_filled .afasta files:   0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "7c61b8e1eaa7441a833886e0c6471b19",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Processing 1H01_A .afasta files:   0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "22bda076c3d7454e930235861d299829",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Processing 1IRK_A .afasta files:   0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "fdae988cf8cd4595ba8571a8d51578b5",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Processing 1CDK_A_filled .afasta files:   0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "2c3c64a979c14b4289cf0b4209d596c9",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Processing 1A9U_A .afasta files:   0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "824e483cc7fb4fcfb330a986479a6882",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Processing 1GAG_A_filled .afasta files:   0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "3d92ce97b3d945f39c3322270e11e5a4",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Processing 1O6L_A_filled .afasta files:   0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "64a085b63af545c79f6ce8cb146d65bf",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Processing 1FPU_A .afasta files:   0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "44df375a7d4f42c88042d7901f18b919",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Processing 1CMK_E_filled .afasta files:   0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "fec1a69ac65b43ce96ccfc73e6924078",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Processing 1L3R_E_filled .afasta files:   0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "fbf68ece4dd043d69cfb14afd526e2ce",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Processing 1BKX_A_filled .afasta files:   0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "99bd348fde5e406e9016b787fe588520",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Processing 1KSW_A .afasta files:   0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "42f61df1b8bc406ab846334f3348a8dc",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Processing 1GZ8_A .afasta files:   0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "6e0df35c75aa4319bfee637822554096",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Processing 1H1P_A_filled .afasta files:   0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "3384d66d239f44aebf3c5e58a19bec50",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Processing 1JOW_B .afasta files:   0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "c3dbdcff0a1142aa867208ad7f997089",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Processing 1JBP_E_filled .afasta files:   0%|          | 0/1 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import pickle\n",
    "import numpy as np\n",
    "from matplotlib import pyplot as plt\n",
    "from sklearn.decomposition import PCA\n",
    "from glob import glob as g\n",
    "import os\n",
    "import matplotlib as mpl\n",
    "from tqdm.notebook import tqdm\n",
    "mpl.rcParams['figure.dpi'] = 300\n",
    "mpl.rcParams.update({'font.size': 8})\n",
    "kind = \"mustang\"\n",
    "make_align_pickle(kind) #create MSA file --> important, substituted .p with .html in alignment class\n",
    "aligned = load_alignments(kind)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "e0a71ccc",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Some helper functions to extract the sequence from the pdb files and to extract the sequence from the alignment files\n",
    "def braf_res():\n",
    "    fp = \"./6UAN_chainD.pdb\"\n",
    "    top = md.load(fp).top\n",
    "    return [res_namer(res) for res in top.residues]\n",
    "\n",
    "\n",
    "def res_namer(res):\n",
    "    return f\"{res.name}-{res.resSeq}\"\n",
    "\n",
    "def fname(fp):\n",
    "    return fp.rsplit(\".\",1)[0].rsplit(\"/\",1)[-1]\n",
    "\n",
    "def make_seg(a):\n",
    "    seq = [t for t in a.aligned if t[0] != \"-\"]\n",
    "    return seq"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "8f3ed920",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 3000x1500 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "\"\"\"\n",
    "Plotting the number of aligned residues over the BRAF BLAST search results to show what are the most conserved residues throughout the alignment\n",
    "\"\"\"\n",
    "from matplotlib.ticker import FuncFormatter\n",
    "seq1mag = len(aligned[0].seq1.replace(\"-\",\"\"))\n",
    "counts = np.zeros(seq1mag)\n",
    "for a in aligned:\n",
    "    segment = make_seg(a)\n",
    "    for i,(b,c) in enumerate(segment):\n",
    "        if c != \"-\":\n",
    "            counts[i] += 1\n",
    "counts =  counts / max(counts)\n",
    "# sns.set_theme(style=\"whitegrid\")\n",
    "fig,ax = plt.subplots(1,figsize=(10,5))\n",
    "x = [*range(len(counts))]\n",
    "ax.set_xticks(x[::5])\n",
    "x_lbl = braf_res()\n",
    "ax.set_xticklabels(x_lbl[::5],rotation=90,fontsize=7)\n",
    "# subtract 11\n",
    "ax.axvspan(36,48, facecolor='g', alpha=0.5)\n",
    "ax.axvspan(92,100, facecolor='c', alpha=0.5)\n",
    "ax.axvspan(144,168, facecolor='r', alpha=0.5)\n",
    "ax.axvspan(177,186, facecolor='y', alpha=0.5)\n",
    "ax.axvspan(204,215, facecolor='pink', alpha=0.8)\n",
    "ax.axvspan(222,240, facecolor='dodgerblue', alpha=0.8)\n",
    "ax.bar(x,counts,linewidth=0.05,width=1)\n",
    "ax.yaxis.set_major_formatter(FuncFormatter(lambda y, _: '{:.0%}'.format(y))) \n",
    "title = ax.set_title(\"Number of MUSTANG aligned residues over the BRAF BLAST search results\")\n",
    "ax1 = plt.xlabel(\"Resiude Name-Number\")\n",
    "ax1 = plt.ylabel(\"Percent matching in structural alignments\")\n",
    "ax.tick_params(length=2,color=\"black\",direction=\"out\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "792d1493",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1 / 38  don't match.\n",
      "Continuing with 37 structures\n",
      "37 structures with an alignment to the D and A\n"
     ]
    }
   ],
   "source": [
    "\"\"\"\n",
    "This cell finds PDBs with DFG and APE motifs aligned\n",
    "\"\"\"\n",
    "from collections import Counter\n",
    "count = 0\n",
    "new_aligned = []\n",
    "bad = []\n",
    "for k,a in enumerate(aligned):\n",
    "    # If doesn't have anything matched totally to DFG and APE discard it\n",
    "    if a.seq1.find(\"DFG\") == -1 or a.seq1.find(\"APE\") == -1:\n",
    "        # print(a.name)\n",
    "        bad.append(aligned)\n",
    "        count += 1\n",
    "    else:\n",
    "        new_aligned.append(a)\n",
    "print(count,\"/\",len(aligned), \" don't match.\")\n",
    "print(f\"Continuing with {len(new_aligned)} structures\")\n",
    "\n",
    "counts  = {}\n",
    "counter = 0\n",
    "new_aligned2 = []\n",
    "aligning_segs = {}\n",
    "lengths = []\n",
    "for a in new_aligned:\n",
    "    DFG_index = a.seq1.find(\"DFG\")\n",
    "    APE_index = a.seq1.find(\"APE\")\n",
    "    dfg = a.seq2[DFG_index]\n",
    "    ape = a.seq2[APE_index]\n",
    "    if dfg != \"-\" and ape != \"-\":\n",
    "        counter+=1\n",
    "        length = APE_index - DFG_index\n",
    "        counts.setdefault(length,[]).append((a.name,DFG_index,APE_index))\n",
    "        new_aligned2.append(a)\n",
    "        aligning_segs.setdefault(a.name,\n",
    "                                 (a.seq2[DFG_index:DFG_index+3],\n",
    "                                  a.seq2[APE_index:APE_index+3],\n",
    "                                  a.seq2[DFG_index:APE_index]))\n",
    "        lengths.append(len(a.seq2[DFG_index:APE_index]))\n",
    "    else:\n",
    "        pass\n",
    "        # print(a.name,dfg,ape)\n",
    "print(f\"{counter} structures with an alignment to the D and A\")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "3d518b58",
   "metadata": {},
   "outputs": [],
   "source": [
    "from glob import glob\n",
    "import mdtraj as md\n",
    "\n",
    "\"\"\"\n",
    "Strips to CAs of activation segment\n",
    "\"\"\"\n",
    "def ifnotmake(dir_path):\n",
    "    if not os.path.isdir(dir_path):\n",
    "        os.makedirs(dir_path)\n",
    "    return dir_path\n",
    "def align_to_actv(pdb,\n",
    "                  alignment,\n",
    "                  align_distance=50):\n",
    "    pdb = md.load(pdb)\n",
    "    DFG_index = alignment.seq1.find(\"DFG\")\n",
    "    APE_index = alignment.seq1.find(\"APE\") + 2 \n",
    "    \n",
    "    '''The following is to make sure that gaps are not counted in the index'''\n",
    "    DFG_index = DFG_index - sum([1 for a in alignment.seq2[:DFG_index] if a == \"-\"])\n",
    "    APE_index = APE_index - sum([1 for a in alignment.seq2[:APE_index] if a == \"-\"])\n",
    "    \n",
    "    actv_range = range(DFG_index, APE_index)\n",
    "    \n",
    "    \n",
    "    atoms_indices = [[atom.index for atom in res.atoms if atom.name == \"CA\"] for res in pdb.top._residues if res.index in actv_range]\n",
    "    atoms_indices = sum(atoms_indices, [])\n",
    "    return pdb.atom_slice(atoms_indices)\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "65927bf7",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "59cb28ab6e20431aa59e42793e4579e6",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "  0%|          | 0/37 [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "'''The following block of code is the same as in the case of align_to_actv version with top_seq_matcher'''\n",
    "names = []\n",
    "rmsds = []\n",
    "target_dir = f\"Results/activation_segments/CA_segments/{kind}/\"\n",
    "ifnotmake(target_dir)\n",
    "# Ensure the target directory exists\n",
    "os.makedirs(target_dir, exist_ok=True)\n",
    "\n",
    "# Get PDB names\n",
    "pdb_dir = \"Results/activation_segments/mustangs/\"\n",
    "fps = glob(pdb_dir+\"/*/*.pdb\")\n",
    "\n",
    "# Process alignments\n",
    "for align_obj in tqdm(new_aligned2):\n",
    "    match_found = False\n",
    "    for fp in fps:\n",
    "        if align_obj.name in fp and \"pdb\" in fp:\n",
    "            match_found = True\n",
    "            break\n",
    "    \n",
    "    if not match_found:\n",
    "        continue\n",
    "    \n",
    "    stripped = align_to_actv(fp, align_obj)\n",
    "    \n",
    "    '''\n",
    "    #Not sure if this is necessary\n",
    "    if stripped.n_residues > 35 or stripped.n_residues < 10:\n",
    "        continue\n",
    "    '''\n",
    "\n",
    "    # Construct the file name and save the stripped pdb\n",
    "    new_name = os.path.join(target_dir, align_obj.name + \".pdb\")\n",
    "    stripped.save(new_name)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "56ef154a",
   "metadata": {},
   "outputs": [],
   "source": [
    "#Code to save only first and last residue in activation segment for further visualization (mainly to double check if mustang alignment makes sense or not)\n",
    "\n",
    "import biobox as bb\n",
    "import glob\n",
    "import numpy as np\n",
    "\n",
    "files = f\"Results/activation_segments/CA_segments/{kind}/*.pdb\"\n",
    "\n",
    "start = []\n",
    "end = []\n",
    "for f in glob.glob(files):\n",
    "\n",
    "    M = bb.Molecule(f)\n",
    "    pts = M.atomselect(\"*\", \"*\", \"CA\")\n",
    "\n",
    "    start.append(pts[0])\n",
    "    end.append(pts[-1])\n",
    "\n",
    "#start = np.array(start)\n",
    "#end = np.array(end)\n",
    "\n",
    "M_start = bb.Structure(p=np.array(start))\n",
    "M_end = bb.Structure(p=np.array(end))\n",
    "\n",
    "M_start.write_pdb(\"start_CA.pdb\")\n",
    "M_end.write_pdb(\"end_CA.pdb\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "14d358bb",
   "metadata": {},
   "outputs": [],
   "source": [
    "#this is right and it works, but renumbers residues meaning that the ones that are missing are ignored\n",
    "\n",
    "from Bio import PDB\n",
    "import numpy as np\n",
    "from scipy.interpolate import interp1d\n",
    "import os\n",
    "import tempfile\n",
    "import mdtraj as md\n",
    "from scipy.interpolate import make_interp_spline\n",
    "def fitting_code(fp_or_traj, save_path):\n",
    "    # Function to read PDB file or trajectory object and get model\n",
    "    def read_structure(input_data):\n",
    "        if isinstance(input_data, str):\n",
    "            # If input is a string, treat it as a file path\n",
    "            parser = PDB.PDBParser(QUIET=True)\n",
    "            structure = parser.get_structure('structure', input_data)\n",
    "        elif isinstance(input_data, md.Trajectory):\n",
    "            # If input is a trajectory, save to temp PDB and read\n",
    "            with tempfile.NamedTemporaryFile(suffix=\".pdb\", delete=False) as tmpfile:\n",
    "                input_data.save(tmpfile.name)\n",
    "                tmpfile.close()\n",
    "                parser = PDB.PDBParser(QUIET=True)\n",
    "                structure = parser.get_structure('structure', tmpfile.name)\n",
    "            os.unlink(tmpfile.name)\n",
    "        else:\n",
    "            raise ValueError(\"Unsupported input type. Provide a file path or md.Trajectory.\")\n",
    "        return structure[0]\n",
    "\n",
    "    # Read the template for CA atoms configuration\n",
    "    template_model = read_structure('template.pdb')\n",
    "    Nnew = len([atom for atom in template_model.get_atoms() if atom.get_id() == 'CA'])\n",
    "\n",
    "    # Read input PDB file or trajectory\n",
    "    my_model = read_structure(fp_or_traj)\n",
    "    atom_list = [atom for atom in my_model.get_atoms() if atom.get_id() == 'CA']\n",
    "    #print(f'shape atom list in input{np.shape(atom_list)}')\n",
    "\n",
    "    n = len(atom_list)\n",
    "    avg = np.array([[atom.coord[0], atom.coord[1], atom.coord[2]] for atom in atom_list])\n",
    "    #print(f'shape atom coordinates in input{np.shape(avg)}')\n",
    "\n",
    "    \n",
    "    dims = ['x', 'y', 'z']\n",
    "    fits = {}\n",
    "    #splines = {}\n",
    "    # Fit cubic interpolation for each axis\n",
    "    for j, dim in enumerate(dims):\n",
    "        #splines[dim] = make_interp_spline(np.arange(n), avg[:, j], k=3)  # 'k=3' for cubic spline \n",
    "        fits[dim] = interp1d(np.arange(n), avg[:, j], kind='cubic', fill_value='extrapolate')\n",
    "    print(avg)\n",
    "    #print(f'fits{fits}')\n",
    "    # Interpolation scales\n",
    "    X = np.arange(0, n-1, 0.1)\n",
    "    #X = np.arange(0, n - 0.2, 0.1)\n",
    "\n",
    "    # Get derivative\n",
    "    dYdX = {dim: np.gradient(fits[dim](X)) for dim in dims}\n",
    "    #evaluated_splines = {dim: np.gradient(splines[dim](X)) for dim in dims}\n",
    "    #print(f'derivative{dYdX}')\n",
    "    # Calculate path length\n",
    "    Y = np.sqrt(sum(np.square(dYdX[dim]) for dim in dims))\n",
    "    #Y = np.sqrt(sum(np.square(evaluated_splines[dim]) for dim in dims))\n",
    "    L = np.trapz(Y, X)\n",
    "\n",
    "    # Create an interpolated arc length\n",
    "    Li = np.linspace(0, L, Nnew)\n",
    "    \n",
    "    # Calculate the arc length for each sampled point\n",
    "    flen = np.array([np.trapz(Y[:ibig], X[:ibig]) for ibig in range(1, len(X))])\n",
    "\n",
    "    pt = np.zeros(Nnew, dtype=int)\n",
    "    for i in range(Nnew):\n",
    "        pt[i] = np.argmin(np.abs(flen - Li[i]))\n",
    "\n",
    "    new_coords = np.array([[fits[dim](X[pt[i]]) for dim in dims] for i in range(Nnew)])\n",
    "    #new_coords = np.array([[splines[dim](X[pt[i]]) for dim in dims] for i in range(Nnew)])\n",
    "    ca_index = 0\n",
    "    for atom in template_model.get_atoms():\n",
    "            if atom.get_id() == 'CA':\n",
    "                atom.set_coord(new_coords[ca_index])\n",
    "                ca_index += 1\n",
    "        \n",
    "        # Now let's write the updated PDB to the file\n",
    "    try:\n",
    "        with open(save_path, \"w\") as file:\n",
    "            io = PDB.PDBIO()\n",
    "            io.set_structure(template_model)\n",
    "            io.save(file)\n",
    "        print(f'Successfully saved the structure to {save_path}')\n",
    "    except Exception as e:\n",
    "        print(f\"Error during file save: {e}\")\n",
    "    pt = np.zeros(Nnew, dtype=int)\n",
    "    for i in range(Nnew):\n",
    "        pt[i] = np.argmin(np.abs(flen - Li[i]))\n",
    "\n",
    "    new_coords = np.array([[fits[dim](X[pt[i]]) for dim in dims] for i in range(Nnew)])\n",
    "    #new_coords = np.array([[splines[dim](X[pt[i]]) for dim in dims] for i in range(Nnew)])\n",
    "    ca_index = 0\n",
    "    for atom in template_model.get_atoms():\n",
    "            if atom.get_id() == 'CA':\n",
    "                atom.set_coord(new_coords[ca_index])\n",
    "                ca_index += 1\n",
    "        \n",
    "        # Now let's write the updated PDB to the file\n",
    "    try:\n",
    "        with open(save_path, \"w\") as file:\n",
    "            io = PDB.PDBIO()\n",
    "            io.set_structure(template_model)\n",
    "            io.save(file)\n",
    "        print(f'Successfully saved the structure to {save_path}')\n",
    "    except Exception as e:\n",
    "        print(f\"Error during file save: {e}\")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "6b5fc504-c470-43da-9076-5d83bac25afc",
   "metadata": {},
   "outputs": [],
   "source": [
    "from glob import glob\n",
    "import os\n",
    "# Start MATLAB\n",
    "#import matlab.engine\n",
    "from tqdm.notebook import tqdm\n",
    "TARGET_DIR = \"Results/fitted_matlab_segments/mustang\"\n",
    "os.makedirs(TARGET_DIR, exist_ok=True)\n",
    "def pdb_id(fp):\n",
    "    return fp.rsplit(\".\",1)[0].rsplit(\"/\",1)[1]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "6fa60c95",
   "metadata": {},
   "outputs": [],
   "source": [
    "PDB_PATH = \"Results/activation_segments/CA_segments/mustang/\"\n",
    "#xyz = md.load(\"/home/marmatt/Documents/projects/BRAF/results/activation_segments/CA_segments/mustang/1A9U_A.pdb\")\n",
    "files = glob(PDB_PATH+\"/*.pdb\")\n",
    "ids = [pdb_id(f) for f in files]\n",
    "print(ids)\n",
    "out = [TARGET_DIR + f\"/{i}.pdb\" for i in ids]\n",
    "#print(files[0],ids[0],out[0])\n",
    "\n",
    "for f, n in tqdm(zip(files,out),total=len(files)):\n",
    "    fitting_code(f,n)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "dc66c173",
   "metadata": {},
   "outputs": [],
   "source": [
    "from matplotlib import pyplot as plt\n",
    "from mpl_toolkits import mplot3d\n",
    "%matplotlib inline\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "import mdtraj as md\n",
    "\n",
    "\n",
    "# Usage Example\n",
    "xyz = md.load(\"Results/activation_segments/CA_segments/mustang/1AD5_A.pdb\")\n",
    "\n",
    "DFG = xyz.top.to_fasta()[0].find(\"DFG\")\n",
    "APE = xyz.top.to_fasta()[0].find(\"APE\")+2\n",
    "atoms = sum([[atom.index for atom in res.atoms if atom.name == \"CA\"] for res in xyz.top._residues[:]],[])\n",
    "\n",
    "coords = xyz.xyz[0,atoms].T\n",
    "x = coords[0]\n",
    "y = coords[1]\n",
    "z = coords[2]\n",
    "new_coords = md.load(\"Results/fitted_matlab_segments/mustang/1AD5_A.pdb\")\n",
    "\n",
    "atoms = sum([[atom.index for atom in res.atoms if atom.name == \"CA\"] for res in new_coords.top._residues[:]],[])\n",
    "new_coords = new_coords.xyz[0,atoms].T\n",
    "xp = new_coords[0]\n",
    "yp = new_coords[1]\n",
    "zp = new_coords[2]\n",
    "\n",
    "fig = plt.figure(figsize=(10,10))\n",
    "ax = plt.axes(projection='3d')\n",
    "ax.plot3D(x,y,z, 'blue',marker=\"o\")\n",
    "ax.plot3D(xp, yp, zp, 'red')\n",
    "plt.tick_params(bottom=False, top=False, labelbottom=False)\n",
    "ax.set_xticks([])\n",
    "ax.set_yticks([])\n",
    "ax.set_zticks([])"
   ]
  }
 ],
 "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": "3.10.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}