def add_atom(self, type):
"""
Adding an atom of given type at the last position in the molecule ordering
:param type:
:return: The index of the added atom
"""
at = Atom(type)
at.SetBoolProp("mutability", True)
self.mol_graph.AddAtom(at)
return at
def test9Validate(self):
vm = rdMolStandardize.RDKitValidation()
mol = Chem.MolFromSmiles("CO(C)C", sanitize=False)
msg = vm.validate(mol)
self.assertEqual(len(msg), 1)
self.assertEqual
("""INFO: [ValenceValidation] Explicit valence for atom # 1 O, 3, is greater than permitted""",
msg[0])
vm2 = rdMolStandardize.MolVSValidation(
[rdMolStandardize.FragmentValidation()])
# with no argument it also works
# vm2 = rdMolStandardize.MolVSValidation()
mol2 = Chem.MolFromSmiles(
"COc1cccc(C=N[N-]C(N)=O)c1[O-].O.O.O.O=[U+2]=O")
msg2 = vm2.validate(mol2)
self.assertEqual(len(msg2), 1)
self.assertEqual
("""INFO: [FragmentValidation] water/hydroxide is present""", msg2[0])
vm3 = rdMolStandardize.MolVSValidation()
mol3 = Chem.MolFromSmiles("C1COCCO1.O=C(NO)NO")
msg3 = vm3.validate(mol3)
self.assertEqual(len(msg3), 2)
self.assertEqual
("""INFO: [FragmentValidation] 1,2-dimethoxyethane is present""",
msg3[0])
self.assertEqual
("""INFO: [FragmentValidation] 1,4-dioxane is present""", msg3[1])
atomic_no = [6, 7, 8]
allowed_atoms = [Atom(i) for i in atomic_no]
vm4 = rdMolStandardize.AllowedAtomsValidation(allowed_atoms)
mol4 = Chem.MolFromSmiles("CC(=O)CF")
msg4 = vm4.validate(mol4)
self.assertEqual(len(msg4), 1)
self.assertEqual
("""INFO: [AllowedAtomsValidation] Atom F is not in allowedAtoms list""",
msg4[0])
atomic_no = [9, 17, 35]
disallowed_atoms = [Atom(i) for i in atomic_no]
vm5 = rdMolStandardize.DisallowedAtomsValidation(disallowed_atoms)
mol5 = Chem.MolFromSmiles("CC(=O)CF")
msg5 = vm4.validate(mol5)
self.assertEqual(len(msg5), 1)
self.assertEqual
("""INFO: [DisallowedAtomsValidation] Atom F is in disallowedAtoms list""",
msg5[0])
msg6 = rdMolStandardize.ValidateSmiles("ClCCCl.c1ccccc1O")
self.assertEqual(len(msg6), 1)
self.assertEqual
("""INFO: [FragmentValidation] 1,2-dichloroethane is present""",
msg6[0])
def try_get_atom_feature(atom: Atom, feat_key: str):
try:
if feat_key == 'is_supernode':
return 0
elif feat_key == 'is_product':
return 0
elif feat_key == 'is_edited':
if atom.HasProp('is_edited') and atom.GetBoolProp('is_edited'):
return 1
return 0
elif feat_key == 'is_reactant':
if atom.HasProp('in_target') and atom.GetBoolProp('in_target'):
return 1
return 0
elif feat_key == 'atomic_num':
return atom.GetAtomicNum()
elif feat_key == 'chiral_tag':
return int(atom.GetChiralTag())
elif feat_key == 'formal_charge':
return atom.GetFormalCharge()
elif feat_key == 'is_aromatic':
return int(atom.GetIsAromatic())
elif feat_key == 'num_explicit_hs':
return atom.GetNumExplicitHs()
else:
raise KeyError(f"Unknown atom feature: {feat_key}")
except RuntimeError as e:
logger.warning(f'Runtime error while try_get_atom_feature: {str(e)}')
return None
def construct_mol(atom_id, A, atomic_num_list):
mol = RWMol()
atomic_num_list = np.array(atomic_num_list, dtype=np.int)
atomic_num = list(filter(lambda x: x > 0, atomic_num_list[atom_id]))
atoms_exist = np.array(list(map(lambda x: x > 0,
atomic_num_list[atom_id]))).astype(np.bool)
atoms_exist = np.expand_dims(atoms_exist, axis=1)
exist_matrix = atoms_exist * atoms_exist.T
for atom in atomic_num:
mol.AddAtom(Atom(int(atom)))
# A (edge_type, num_node, num_node)
adj = np.argmax(A, axis=0)
adj = adj[exist_matrix].reshape(len(atomic_num), len(atomic_num))
for i, j in zip(*np.nonzero(adj)):
if i > j:
mol.AddBond(int(i), int(j), adj_to_bond_type[adj[i, j]])
return mol
def validate_allowed_atoms(mol, atomlist):
"""Validates by a list of allowed atoms.
Parameters
----------
mol: rdkit.Chem.Mol
A molecule.
atomlist: list of int
The atomic number of the allowed atoms.
Returns
-------
message: str
Error Message with every atom that is not on atomlist.
"""
if len(atomlist) == 0:
return
else:
allowed_atoms = [Atom(i) for i in atomlist]
vm = rdMolStandardize.AllowedAtomsValidation(allowed_atoms)
message = vm.validate(mol)
return message or None