def read_standardise(smi) -> Chem.rdchem.Mol:
null_mol = Chem.MolFromSmiles('')
try:
m = Chem.MolFromSmiles(smi)
assert m is not None
except:
warnings.warn('SMILES: {} is invalid'.format(smi))
return null_mol
m = standardizer.standardize_mol(m)
m = standardizer.get_parent_mol(m)[0]
return m
def standardize_smiles(smiles):
try:
# Convert to RDKit mol
m = Chem.MolFromSmiles(smiles)
# Neutralize and separate from counterion
m = standardizer.get_parent_mol(m, neutralize=True, check_exclusion=True, verbose=False)[0]
# Standardize representation
m = standardizer.standardize_mol(m, check_exclusion=True)
# Return RDKit canonical SMILES
# remove stereochemistry information to avoid some activity cliffs
return Chem.MolToSmiles(Chem.MolFromSmiles(Chem.MolToSmiles(m, isomericSmiles=False)))
except:
return np.nan
def check_errors(self, smi: str) -> str:
"""
This function processes the SMILES in order to canonicalize it and detect any errors.
If errors are detected, the returned SMILES is not sanitized.
:param smi: SMILES string of the compound
:return final_smi: canonicalized SMILES
"""
try:
final_smi = standardizer.standardize_mol(smi)
except:
final_smi = smi
final_smi = Chem.MolToSmiles(final_smi)
final_smi = self.salt_remover(final_smi)
return final_smi
def __call__(self, mol):
if chembl_standardizer.exclude_flag(mol,
includeRDKitSanitization=False):
raise StandardizationError(
None,
f'ChEMBL standardizer set the exclusion flag for molecule: {Chem.MolToSmiles(mol)}'
)
# just for outputs
try:
smiles = Chem.MolToSmiles(mol)
except Exception as exp:
raise StandardizationError(
exp,
f'An exception occurred while getting the SMILES for molecule: {mol}'
)
try:
mol = chembl_standardizer.standardize_mol(mol,
check_exclusion=False)
except Exception as exp:
raise StandardizationError(
exp,
f'An exception occurred while standardizing molecule: {smiles}'
)
try:
mol, _ = chembl_standardizer.get_parent_mol(mol,
check_exclusion=False,
verbose=True,
neutralize=True)
except Exception as exp:
raise StandardizationError(
exp,
f'An exception occurred while getting the parent molecule of: {smiles}'
)
return mol
# get all csvs from folders
received_csv_files = [
f for f in dir_path.glob("**/*.csv")
if "all_received_mols.csv" not in str(f)
]
smiles_dict = {}
for csv_file in received_csv_files:
try:
received_df = pd.read_csv(csv_file)
received_df["SMILES"] = received_df["SMILES"].apply(
lambda x: Chem.MolToSmiles(
Chem.MolFromSmiles(
Chem.MolToSmiles(
standardizer.standardize_mol(
standardizer.get_parent_mol(Chem.MolFromSmiles(x))[
0])))))
received_smi = list(received_df["SMILES"])
for smi in received_smi:
smiles_dict[smi] = str(csv_file).split("/")[-1]
except Exception as e:
print(f"FAILED ON {csv_file}")
print(e)
pass
# write out final csv
all_smiles = list(smiles_dict.keys())
all_shipments = [smiles_dict[x] for x in all_smiles]
def strip_and_standardize_smi(smi):
return Chem.MolToSmiles(
Chem.MolFromSmiles(
Chem.MolToSmiles(
standardizer.standardize_mol(
standardizer.get_parent_mol(Chem.MolFromSmiles(smi))[0]))))