def predict_outcome(self, list_of_mols, k=1):
"""
Using a predictor, produce top-k most likely reactions
Params:
:list_of_mols: list of reactants and reagents (both of class Molecule)
(former contribute atoms, latter don't)
"""
react = ".".join([m.smiles for m in list_of_mols])
(react, bond_preds, bond_scores,
cur_att_score) = self.directcorefinder.predict(react)
#---> TODO: add input check here: some molecules seem to be 'unparseable' <---#
# this might be a problem of Rexgen, though
outcomes = self.directcandranker.predict(react, bond_preds,
bond_scores)
res = []
for out in outcomes[:k]:
smiles = out["smiles"][0]
mol = Molecule(smiles)
mol.set_synthesis(list_of_mols)
res.append(mol)
return res
def get_chembl(n_mols=None, as_mols=True):
""" Return list of SMILES """
path = "./datasets/ChEMBL.txt"
with open(path, "r") as f:
if n_mols is None:
res = [line.strip() for line in f]
else:
res = [f.readline().strip() for _ in range(n_mols)]
return [Molecule(smile) for smile in res]
def test_sas(self):
sas_func = lambda mol: calculateSAScore(Chem.MolFromSmiles(mol.smiles))
print(sas_func(Molecule("CC")))
test_pool = [
"CC", "O=C=O", "C#N", "CCN(CC)CC", "CC(=O)O", "C1CCCCC1",
"c1ccccc1"
]
test_pool = [Molecule(smiles) for smiles in test_pool]
exp = RandomExplorer(sas_func, initial_pool=test_pool)
print("Starting SA score optimization")
t0 = time()
exp.evolve(10)
#check
print("Completed SA score optimization, time elapsed: %.3fs" %
(time() - t0))
print(exp.pool)
top = exp.get_best(1)[0]
print(top.get_synthesis_path())
def test_len(self):
dummy_func = lambda mol: len(mol.smiles)
test_pool = [
"CC", "O=C=O", "C#N", "CCN(CC)CC", "CC(=O)O", "C1CCCCC1",
"c1ccccc1"
]
test_pool = [Molecule(smiles) for smiles in test_pool]
exp = RandomExplorer(dummy_func, initial_pool=test_pool)
print("Starting len of SMILES optimization")
exp.evolve(2)
#check
print(exp.pool)
def ga_optimise_from_args(func, max_capital):
# the func may accept iterable or a single Molecule
mol = Molecule(smiles="c1cc(OCCCN2CCCCC2)ccc1CN1CCC2(CC1)OCCO2")
try:
func(mol)
func_ = func
except Exception as e:
# print("Failed,", e)
func_ = lambda m: func([m])
explorer = RandomExplorer(func_)
explorer.evolve(max_capital)
top = explorer.get_best(k=1)
val = func(top)
return top, val
def test_chembl(self):
"""
Problem with fixed-prop testing:
Almost all of the results (<10% for init_pool of 50) seem to be outside of the database,
and even less for smaller pool. Hence cannot get its score for testing;
setting them to zero leads to slow exploration.
"""
pool_all, dd = chembl_prop_exp()
# loading with mol conversions takes 8 minutes
# pool_all = [Molecule(smiles, conv_enabled=True) for smiles in tqdm(pool_all[:10000])]
pool_all = [
Molecule(smiles, conv_enabled=False) for smiles in pool_all
]
start_pool = list(np.random.choice(pool_all, size=100, replace=False))
def print_props(pool):
props = [dd[mol.smiles] for mol in pool]
print("Props of pool", len(pool), np.min(props), np.mean(props),
np.max(props))
print_props(pool_all)
print_props(start_pool)
func = lambda mol: dd[mol.smiles]
exp = RandomExplorer(func, initial_pool=start_pool)
print("Starting ChEMBL score 1 optimization")
t0 = time()
exp.evolve(30)
print("Completed ChEMBL score 1 optimization, time elapsed: %.3fs" %
(time() - t0))
# print(exp.pool)
top = exp.get_best(1)[0]
print(top.get_synthesis_path())
print("Best achieved score: %.3f" % func(top))
props = [dd[mol.smiles] for mol in pool_all]
print("Best possible score: %.3f" % np.max(props))
def test_wl_kernel(self):
mols = [Molecule(S1), Molecule(S2), Molecule(S3)]
print(mol_kernels.compute_wl_kernel(mols))
def test_edgehist_kernel(self):
mols = [Molecule(S1), Molecule(S2)]
print(mol_kernels.compute_edgehist_kernel(mols))
def test_conversions(self):
mol = Molecule(S1)
graph = mol_kernels.mol2graph_igraph(mol)
print(graph)