Update REDOX Interference_ preprocessing script.py

REDOX Interference_ preprocessing script.py

@@ -1,3 +1,5 @@
+#1. Import modules
+
 pip install rdkit
 pip install molvs
 import pandas as pd
@@ -9,17 +11,19 @@ from rdkit import Chem
 standardizer = molvs.Standardizer()
 fragment_remover = molvs.fragment.FragmentRemover()
 
+# 2. Convert the SDF file from the original paper into data frame
+# Before running the code, please download SDF files from the original paper
+# https://pubs.acs.org/doi/10.1021/acs.jmedchem.3c00482
+
 from rdkit.Chem import PandasTools
 sdfFile = 'Redox_training_set_curated.sdf'
 dataframe = PandasTools.LoadSDF(sdfFile)
-
-dataframe.to_csv('Nano Luciferase.csv', index=False)
-
+dataframe.to_csv('redox.csv', index=False)
 df = pd.read_csv('redox.csv')
 
-# Some of the 'Raw_SMILES' rows contain TWO smiles separated by ;
-# These cause smiles parse error (which means they cannot be read)
-# So I separated the smiles
+# 3. Resolve SMILES parse error
+# Some of the 'Raw_SMILES' rows contain TWO SMILES separated by ';'' and, they cause SMILES parse error (which means they cannot be read)
+# So we separated the SMILES and renamed the columns
 
 df.rename(columns = {'PUBCHEM_EXT_DATASOURCE_REGID': 'REGID_1'}, inplace = True)
 df.rename(columns = {'Other REGIDs': 'REGID_2'}, inplace = True)
@@ -34,6 +38,8 @@ df['AC50_uM_2'] = df['AC50_uM'].str.split(';').str[1]
 df['AC50_uM'] = df['AC50_uM'].str.split(';').str[0]
 df.rename(columns = {'AC50_uM': 'AC50_uM_1'}, inplace = True)
 
+# 4. Sanitize with MolVS and print problems
+
 df['X_1'] = [ \
     rdkit.Chem.MolToSmiles(
         fragment_remover.remove(
@@ -56,8 +62,12 @@ def process_smiles(smiles):
 
 df['X_2'] = df['SMILES_2'].apply(process_smiles)
 
+# 5. Rename the columns
+
 df.rename(columns={'X_1' : 'newSMILES_1', 'X_2' : 'newSMILES_2'}, inplace = True)
 
+# 6. Create a file with sanitized SMILES
+
 df[['REGID_1',
  'REGID_2',
  'newSMILES_1',