Compare two SMILES or SD files to identify duplicate structures. Duplicates are identified by comparing InChI keys

dupe.py

#!/usr/bin/env python

import sys
import os
from rdkit import Chem


def molecule_supplier_from_name(input_file_name):
    ext = os.path.splitext(input_file_name)[-1]
    if ext == ".smi":
        suppl = Chem.SmilesMolSupplier(input_file_name)
    elif ext == ".sdf":
        suppl = Chem.SDMolSupplier(input_file_name)
    else:
        print("%s is not a valid molecule extension" % ext)
        sys.exit(0)
    return suppl


def build_inchi_dict(infile_name):
    suppl = molecule_supplier_from_name(infile_name)
    if suppl is None:
        print("Error reading %s, extension not recognized" % infile_name)
        sys.exit(1)
    inchi_dict = {}
    for mol in suppl:
        if mol:
            name = mol.GetProp("_Name")
            inchi_key = Chem.MolToInchiKey(mol)
            inchi_dict[inchi_key] = name
    return inchi_dict


def check_infile(infile_name, inchi_dict):
    suppl = molecule_supplier_from_name(infile_name)
    for mol in suppl:
        if mol:
            name = mol.GetProp("_Name")
            smiles = Chem.MolToSmiles(mol, isomericSmiles=True)
            inchi = Chem.MolToInchiKey(mol)
            match = inchi_dict.get(inchi) or "NO_MATCH"
            print(smiles, name, match)


if __name__ == "__main__":
    usage = """%s reference.[sdf|smi] query.[sdf|smi]
Program returns the matching molecule name from reference or NO_MATCH 
for each molecule in query 
    """ % sys.argv[0]
    if len(sys.argv) != 3:
        print(usage, file=sys.stderr)
        sys.exit(0)
    ref_dict = build_inchi_dict(sys.argv[1])
    check_infile(sys.argv[2], ref_dict)