def set_reaction(self,product_smiles):
try:
product_mol = SM(product_smiles)
prod_num_rings = CalcNumRings(product_mol)
except:
print("error in the product",product_smiles)
return []
try:
reactant_list = self.rxn.RunReactants([product_mol])
except:
print("Reaction failed")
print(self.reaction_name,self.smarts,product_smiles)
exit()
approved_reactants = []
for reactant_mol in reactant_list:
#condition 1 - conserved ring count
try:
[Chem.SanitizeMol(r) for r in reactant_mol]
if np.sum([CalcNumRings(r) for r in reactant_mol]) - prod_num_rings == self.ring_change_count:
approved_reactants.append(reactant_mol)
except:
print("could not sanitize ",product_smiles,".".join([MS(r) for r in reactant_mol]))
return approved_reactants
def convert_to_featmorgan(SMILES):
mol = MS(SMILES)
morgan = Chem.GetMorganFingerprintAsBitVect(mol,
2,
nBits=1024,
useFeatures=True)
morgan_float_list = list(
np.asarray(list(morgan.ToBitString()), dtype=float))
return morgan_float_list
def decompose(self, mol_id, smiles, reaction_func):
#perform the decomposition reaction
reactant_list = reaction_func.run_reaction(smiles)
output_list = []
for reactant_mol in reactant_list:
reactant_pair = []
for mol in reactant_mol:
#converts reactant rdkit mol objects to smiles
reactant_pair.append(MS(mol))
#makes sure we don't log duplicates of the decomposition reaction
if reactant_pair not in output_list:
output_list.append(
str(mol_id) + "|" + smiles + "|" +
reaction_func.reaction_name + "|" +
".".join(reactant_pair))
return output_list
try:
mol_desc = convert_to_morgan(mol)
dist_val1 = tanimoto_dist(mol_desc, template1)
dist_val2 = tanimoto_dist(mol_desc, template2)
ave_dist = np.average([dist_val1, dist_val2])
zinc_dist_list.append(ave_dist)
except:
#desc_func(mol)
print("skipped smiles: ")
pass
return zinc_dist_list
template_file = open(args['template_file'], "r")
template_desc_list = [
convert_to_morgan(MS(template.strip("\n"))) for template in template_file
]
template_file.close()
print("inputs", zinc_split_index, split_num)
os.system("echo calc_time > zinc_" + str(zinc_split_index) + ".out")
#first calculation
sorted_dist = []
sorted_index = []
sorted_file_label = []
for zf_index, zf in enumerate(
np.array_split(zinc_files, split_num)[zinc_split_index]):
st = time.time()
def convert_to_tor(SMILES):
mol = MS(SMILES)
desc_val = Tor(mol)
desc_val_float_list = list(
np.asarray(list(desc_val.ToBitString()), dtype=float))
return desc_val_float_list
def convert_to_atompair(SMILES):
mol = MS(SMILES)
atom_pair = AtomPair(mol)
atom_pair_float_list = list(
np.asarray(list(atom_pair.ToBitString()), dtype=float))
return atom_pair_float_list
def convert_to_clogp(SMILES):
mol = MS(SMILES)
logp = MolLogP(mol)
return logp
def convert_to_MW(SMILES):
mol = MS(SMILES)
MW = Chem.Descriptors.ExactMolWt(mol)
return MW
def convert_to_maccs(SMILES):
mol = MS(SMILES)
maccs = Chem.GetMACCSKeysFingerprint(mol)
maccs_float_list = list(np.asarray(list(maccs.ToBitString()), dtype=float))
return maccs_float_list
def convert_to_num_rings(SMILES):
mol = MS(SMILES)
val = rdkit.Chem.rdMolDescriptors.CalcNumRings(mol)
return val
def convert_to_num_aromatic_hetero(SMILES):
mol = MS(SMILES)
val = rdkit.Chem.rdMolDescriptors.CalcNumAromaticHeterocycles(mol)
return val
def convert_to_HBD(SMILES):
mol = MS(SMILES)
val = rdkit.Chem.rdMolDescriptors.CalcNumHBD(mol)
return val
def convert_to_cats(SMILES):
mol = MS(SMILES)
cats_desc = cats.getCATs2D(mol)
return cats_desc
def convert_to_rdkit(SMILES):
mol = MS(SMILES)
desc_val = RDKFingerprint(mol)
desc_val_float_list = list(
np.asarray(list(desc_val.ToBitString()), dtype=float))
return desc_val_float_list