def sanitize(population,scores,population_size,prune_population):
if prune_population:
smiles_list = []
population_tuples = []
for score, string in zip(scores,population):
canonical_smiles = Chem.MolToSmiles(co.string2mol(string))
if canonical_smiles not in smiles_list:
smiles_list.append(canonical_smiles)
population_tuples.append((score, string))
else:
population_tuples = list(zip(scores,population))
population_tuples = sorted(population_tuples, key=lambda x: x[0], reverse=True)[:population_size]
new_population = [t[1] for t in population_tuples]
new_scores = [t[0] for t in population_tuples]
return new_population, new_scores
def mutate(mol,mutation_rate):
if random.random() > mutation_rate:
return mol
Chem.Kekulize(mol,clearAromaticFlags=True)
child = stco.mol2string(mol)
symbols = get_symbols()
for i in range(50):
random_number = random.random()
mutated_gene = random.randint(0, len(child) - 1)
random_symbol_number = random.randint(0, len(symbols)-1)
new_child = list(child)
random_number = random.random()
new_child[mutated_gene] = symbols[random_symbol_number]
new_child_mol = stco.string2mol(new_child)
#print(child_smiles,Chem.MolToSmiles(child_mol),child_mol,co.mol_OK(child_mol))
if co.mol_OK(new_child_mol):
return new_child_mol
return mol
def logP_max(string, dummy):
mol = co.string2mol(string)
score = sc.logP_score(mol)
return max(0, score)
def rediscovery(string, args):
mol = co.string2mol(string)
score = sc.rediscovery(mol, args)
return score
def absorbance_target(string, args):
mol = co.string2mol(string)
score = sc.absorbance_target(mol, args)
return score
def logP_target(string, args):
mol = co.string2mol(string)
score = sc.logP_target(mol, args)
return score