Add files via upload

cns_mpo_csv_to_df.py

@@ -0,0 +1,148 @@
+import pandas as pd
+from rdkit import Chem
+from rdkit.Chem import Descriptors, Crippen, rdMolDescriptors
+from math import log10
+
+
+class CNS_MPO_csv_to_df:
+    def __init__(self, csv_file):
+        self.csv_file = csv_file
+        self._df = None
+        self.calculate()
+
+    def read_csv(self):
+        df = pd.read_csv(self.csv_file, sep=";")
+        df["pKa"] = df["pKa"].str.replace(",", ".").astype(float)
+        return df
+
+    def clogD(self, logP, pKa, pH=7.4):
+        return logP - log10(1 + 10 ** (pH - pKa))
+
+    def csv_file_preparation(self):
+        dictionary = {"MW": [], "LogP": [], "HBD": [], "TPSA": []}
+
+        for cpd in self._df["Smiles"]:
+            molecule = Chem.MolFromSmiles(cpd)
+            mol_mw = Descriptors.MolWt(molecule)
+            mol_logp = Crippen.MolLogP(molecule)
+            mol_hbd = rdMolDescriptors.CalcNumHBD(molecule)
+            mol_tpsa = Descriptors.TPSA(molecule)
+
+            dictionary["MW"].append(mol_mw)
+            dictionary["LogP"].append(mol_logp)
+            dictionary["HBD"].append(mol_hbd)
+            dictionary["TPSA"].append(mol_tpsa)
+
+        df_descriptors = pd.DataFrame(dictionary)
+        df_descriptors["pKa"] = self._df["pKa"]
+        df_descriptors["LogD"] = df_descriptors.apply(
+            lambda x: self.clogD(x["LogP"], x["pKa"]), axis=1
+        )
+        return df_descriptors
+
+    def mw_score_func(self, mw):
+        if mw <= 360:
+            return 1
+        elif 360 < mw <= 500:
+            return -0.005 * mw + 2.5
+        else:
+            return 0
+
+    def logp_score_func(self, logp):
+        if logp <= 3:
+            return 1
+        elif 3 < logp <= 5:
+            return -0.5 * logp + 2.5
+        else:
+            return 0
+
+    def logd_score_func(self, logd):
+        if logd <= 2:
+            return 1
+        elif 2 < logd <= 4:
+            return -0.5 * logd + 2
+        else:
+            return 0
+
+    def pka_score_func(self, pka):
+        if pka <= 8:
+            return 1
+        elif 8 < pka <= 10:
+            return -0.5 * pka + 5
+        else:
+            return 0
+
+    def tpsa_score_func(self, tpsa):
+        if 40 <= tpsa <= 90:
+            return 1
+        elif 90 < tpsa <= 120:
+            return -0.0333 * tpsa + 4
+        elif 20 <= tpsa < 40:
+            return 0.05 * tpsa - 1
+        else:
+            return 0
+
+    def hbd_score_func(self, hbd):
+        if hbd == 0:
+            return 1
+        elif hbd == 1:
+            return 0.75
+        elif hbd == 2:
+            return 0.5
+        elif hbd == 3:
+            return 0.25
+        else:
+            return 0
+
+    def calcCNS_MPO(self):
+        df_descriptors = self.csv_file_preparation()
+        df_descriptors["MW_score"] = df_descriptors["MW"].apply(
+            self.mw_score_func
+        )
+        df_descriptors["LogP_score"] = df_descriptors["LogP"].apply(
+            self.logp_score_func
+        )
+        df_descriptors["LogD_score"] = df_descriptors["LogD"].apply(
+            self.logd_score_func
+        )
+        df_descriptors["pKa_score"] = df_descriptors["pKa"].apply(
+            self.pka_score_func
+        )
+        df_descriptors["TPSA_score"] = df_descriptors["TPSA"].apply(
+            self.tpsa_score_func
+        )
+        df_descriptors["HBD_score"] = df_descriptors["HBD"].apply(
+            self.hbd_score_func
+        )
+
+        df_descriptors["CNS_MPO"] = (
+            df_descriptors["MW_score"]
+            + df_descriptors["LogP_score"]
+            + df_descriptors["LogD_score"]
+            + df_descriptors["pKa_score"]
+            + df_descriptors["TPSA_score"]
+            + df_descriptors["HBD_score"]
+        )
+        df_descriptors["Id"] = self._df["Id"]
+
+        return df_descriptors[
+            ["Id", "MW", "LogP", "LogD", "pKa", "TPSA", "HBD", "CNS_MPO"]
+        ]
+
+    def calculate(self):
+        self._df = self.read_csv()
+        self._df = self.calcCNS_MPO()
+
+    def __getitem__(self, key):
+        if isinstance(key, (int, slice)):
+            return self._df.iloc[key]
+        elif isinstance(key, str):
+            return self._df[key]
+        else:
+            raise KeyError(f"Unsupported key type: {type(key)}")
+
+    def __repr__(self):
+        return repr(self._df)
+
+    def __str__(self):
+        return str(self._df)

cns_mpo_single_molecule.py

@@ -0,0 +1,166 @@
+import pandas as pd
+from rdkit import Chem
+from rdkit.Chem import Descriptors, Crippen, rdMolDescriptors
+from math import log10
+
+
+class CNS_MPO_single_molecule:
+    def __init__(self, smiles_list, pKa_list):
+        if len(smiles_list) != len(pKa_list):
+            raise ValueError(
+                "Length of smiles_list must be equal to length of pKa_list"
+            )
+        self.smiles_list = smiles_list
+        self.pKa_list = pKa_list
+        self._df = None
+        self.calculate()
+
+    def clogD(self, logP, pKa, pH=7.4):
+        return logP - log10(1 + 10 ** (pH - pKa))
+
+    def csv_file_preparation(self):
+        dictionary = {"MW": [], "LogP": [], "HBD": [], "TPSA": []}
+
+        for cpd in self.smiles_list:
+            molecule = Chem.MolFromSmiles(cpd)
+            if molecule is None:
+                # Handle invalid SMILES
+                dictionary["MW"].append(None)
+                dictionary["LogP"].append(None)
+                dictionary["HBD"].append(None)
+                dictionary["TPSA"].append(None)
+                continue
+            mol_mw = Descriptors.MolWt(molecule)
+            mol_logp = Crippen.MolLogP(molecule)
+            mol_hbd = rdMolDescriptors.CalcNumHBD(molecule)
+            mol_tpsa = Descriptors.TPSA(molecule)
+
+            dictionary["MW"].append(mol_mw)
+            dictionary["LogP"].append(mol_logp)
+            dictionary["HBD"].append(mol_hbd)
+            dictionary["TPSA"].append(mol_tpsa)
+
+        df_descriptors = pd.DataFrame(dictionary)
+        df_descriptors["pKa"] = self.pKa_list
+        df_descriptors["LogD"] = df_descriptors.apply(
+            lambda x: self.clogD(x["LogP"], x["pKa"]), axis=1
+        )
+        return df_descriptors
+
+    def mw_score_func(self, mw):
+        if mw is None:
+            return 0
+        if mw <= 360:
+            return 1
+        elif 360 < mw <= 500:
+            return -0.005 * mw + 2.5
+        else:
+            return 0
+
+    def logp_score_func(self, logp):
+        if logp is None:
+            return 0
+        if logp <= 3:
+            return 1
+        elif 3 < logp <= 5:
+            return -0.5 * logp + 2.5
+        else:
+            return 0
+
+    def logd_score_func(self, logd):
+        if logd is None:
+            return 0
+        if logd <= 2:
+            return 1
+        elif 2 < logd <= 4:
+            return -0.5 * logd + 2
+        else:
+            return 0
+
+    def pka_score_func(self, pka):
+        if pka is None:
+            return 0
+        if pka <= 8:
+            return 1
+        elif 8 < pka <= 10:
+            return -0.5 * pka + 5
+        else:
+            return 0
+
+    def tpsa_score_func(self, tpsa):
+        if tpsa is None:
+            return 0
+        if 40 <= tpsa <= 90:
+            return 1
+        elif 90 < tpsa <= 120:
+            return -0.0333 * tpsa + 4
+        elif 20 <= tpsa < 40:
+            return 0.05 * tpsa - 1
+        else:
+            return 0
+
+    def hbd_score_func(self, hbd):
+        if hbd is None:
+            return 0
+        if hbd == 0:
+            return 1
+        elif hbd == 1:
+            return 0.75
+        elif hbd == 2:
+            return 0.5
+        elif hbd == 3:
+            return 0.25
+        else:
+            return 0
+
+    def calcCNS_MPO(self):
+        df_descriptors = self.csv_file_preparation()
+        df_descriptors["MW_score"] = df_descriptors["MW"].apply(
+            self.mw_score_func
+        )
+        df_descriptors["LogP_score"] = df_descriptors["LogP"].apply(
+            self.logp_score_func
+        )
+        df_descriptors["LogD_score"] = df_descriptors["LogD"].apply(
+            self.logd_score_func
+        )
+        df_descriptors["pKa_score"] = df_descriptors["pKa"].apply(
+            self.pka_score_func
+        )
+        df_descriptors["TPSA_score"] = df_descriptors["TPSA"].apply(
+            self.tpsa_score_func
+        )
+        df_descriptors["HBD_score"] = df_descriptors["HBD"].apply(
+            self.hbd_score_func
+        )
+
+        df_descriptors["CNS_MPO"] = (
+            df_descriptors["MW_score"]
+            + df_descriptors["LogP_score"]
+            + df_descriptors["LogD_score"]
+            + df_descriptors["pKa_score"]
+            + df_descriptors["TPSA_score"]
+            + df_descriptors["HBD_score"]
+        )
+
+        # Return only the specified columns
+        return df_descriptors[
+            ["MW", "LogP", "HBD", "TPSA", "pKa", "LogD", "CNS_MPO"]
+        ]
+
+    def calculate(self):
+        self._df = self.calcCNS_MPO()
+
+    def __getitem__(self, key):
+        if isinstance(key, (int, slice)):
+            return self._df.iloc[key]
+        elif isinstance(key, str):
+            return self._df[key]
+        else:
+            raise KeyError(f"Unsupported key type: {type(key)}")
+
+    def __repr__(self):
+        return repr(self._df)
+
+    def __str__(self):
+        return str(self._df)

csv_file_instruction.txt

@@ -0,0 +1,2 @@
+It is important that in the CSV file each column is located in the next cell, and not after a comma.
+Required separator is semicolon (";"), not comma (",").

csv_file_with_example_molecules.csv

@@ -0,0 +1,7 @@
+Id;Smiles;pKa
+telmisartan;CCCC1=NC2=C(N1CC3=CC=C(C4=CC=CC=C4C(O)=O)C=C3)C=C(C5=NC6=CC=CC=C6N5C)C=C2C;3,581
+ibuprofen;CC(CC1=CC=C(C(C(O)=O)C)C=C1)C;4,366
+sulfasalazine;OC(C1=CC(/N=N/C2=CC=C(S(=O)(NC3=NC=CC=C3)=O)C=C2)=CC=C1O)=O;2,715
+risperidone;O=C1C(CCN2CCC(CC2)C3=NOC4=C3C=CC(F)=C4)=C(N=C5N1CCCC5)C;8,765
+paroxetine;C1CNCC(C1C2=CC=C(C=C2)F)COC3=CC4=C(C=C3)OCO4;9,365
+methylophenidate;O=C(OC)C(C1CCCCN1)C2=CC=CC=C2;8,375

example_of_use.py

@@ -0,0 +1,19 @@
+# Example of CNS_MPO_single_molecule() use
+x = CNS_MPO_single_molecule(
+    smiles_list=[
+        "C(C=1CCN(C5)CCC(C5)c(n4)c(c3o4)ccc(c3)F)(=O)N(C2)C(CCC2)=NC1C",
+        "C1CNCC(C1C2=CC=C(C=C2)F)COC3=CC4=C(C=C3)OCO4",
+    ],
+    pKa_list=[8.765, 9.365],
+)
+print(x)
+
+
+# Example of CNS_MPO_csv_to_df() use
+filepath = 'path_to_CSV_files'
+x = CNS_MPO_csv_to_df(csv_file=filepath)
+print(x[1:3][['Id','MW']])  #1 user can specify columns and rows with slicing, which can enable him to get desired output
+print(x[1:3]['Id'])  #2
+print(x[1:3])  #3
+print(x)  #4 but also user can get whole object
+