Read compounds from multiple SDFs and write calculations to a single SDF file

sdf-read-write.py

# necessary imports
import pandas as pd
from rdkit import Chem
from rdkit.Chem import PandasTools
from rdkit.Chem import Crippen
import glob
import pandas as pd

# Dataframe to write calculations of each compounds
compounds = pd.DataFrame(columns=['ID', 'Smiles', 'Molecular_Formula', 'Molecular_Weight', 'H_Bond_Acceptors',
                                   'H_Bond_Donors', 'Molar_Refractivity', 'TPSA'])

# skip duplicates
checkDuplicateSmiles = dict()

i = 0 # index counter
for file in glob.iglob('./**//*.sdf', recursive=True): # data is saved in subdirectories of current directory
    sdf = Chem.SDMolSupplier(file) # read sdf
    for mol in sdf:
        smiles = Chem.MolToSmiles(mol) # get smiles
        if checkDuplicateSmiles.get(smiles, -1) == -1: # if this is not already processed
            checkDuplicateSmiles[smiles] = True # mark it as processed
            i +=1 # index counter
            molecular_formula = Chem.rdMolDescriptors.CalcMolFormula(mol) # formula
            molecular_weight = Chem.rdMolDescriptors.CalcExactMolWt(mol) # weight
            hba = Chem.rdMolDescriptors.CalcNumHBA(mol) # h bond acceptor
            hbd = Chem.rdMolDescriptors.CalcNumHBD(mol) # h bond donor
            molar_refractivity = Chem.Crippen.MolMR(mol) # molar refractivity
            tpsa = Chem.rdMolDescriptors.CalcTPSA(mol) # tpsa
            compounds = compounds.append({ # write this row to dataframe
                'ID': 'Phytochem_' + str(i).zfill(5),
                'Smiles': smiles,
                'Molecular_Formula': molecular_formula,
                'Molecular_Weight': molecular_weight,
                'H_Bond_Acceptors': hba,
                'H_Bond_Donors': hbd,
                'Molar_Refractivity': molar_refractivity,
                'TPSA': tpsa
            }, ignore_index=True)
            
# add molecule column
PandasTools.AddMoleculeColumnToFrame(compounds,'Smiles','ROMol',includeFingerprints=True)
# now write to a single sdf
PandasTools.WriteSDF(compounds, 'new_sdf.sdf', molColName='ROMol', idName='ID', properties=list(compounds.columns))