def create_augmented_layers(mol):
"""
The indices in the string refer to the atom indices in the molecule, according to the atom order
obtained by sorting the atoms using the InChI canonicalization algorithm.
First a deep copy is created of the original molecule and hydrogen atoms are removed from the molecule.
Next, the molecule is converted into an InChI string, and the auxiliary information of the inchification
procedure is retrieved.
The N-layer is parsed and used to sort the atoms of the original order according
to the order in the InChI. In case, the molecule contains atoms that cannot be distinguished
with the InChI algorithm ('equivalent atoms'), the position of the unpaired electrons is changed
as to ensure the atoms with the lowest indices are used to compose the string.
"""
if mol.getRadicalCount() == 0 and not has_unexpected_lone_pairs(mol):
return None, None
elif mol.getFormula() == 'H':
return inchiutil.U_LAYER_PREFIX + '1', None
else:
molcopy = mol.copy(deep=True)
hydrogens = filter(lambda at: at.number == 1, molcopy.atoms)
[molcopy.removeAtom(h) for h in hydrogens]
rdkitmol = toRDKitMol(molcopy)
_, auxinfo = Chem.MolToInchiAndAuxInfo(
rdkitmol, options='-SNon') # suppress stereo warnings
# extract the atom numbers from N-layer of auxiliary info:
atom_indices = inchiutil.parse_N_layer(auxinfo)
atom_indices = [
atom_indices.index(i + 1) for i, atom in enumerate(molcopy.atoms)
]
# sort the atoms based on the order of the atom indices
molcopy.atoms = [
x for (y, x) in sorted(zip(atom_indices, molcopy.atoms),
key=lambda pair: pair[0])
]
ulayer = create_U_layer(molcopy, auxinfo)
player = create_P_layer(molcopy, auxinfo)
return ulayer, player
def create_augmented_layers(mol):
"""
The indices in the string refer to the atom indices in the molecule, according to the atom order
obtained by sorting the atoms using the InChI canonicalization algorithm.
First a deep copy is created of the original molecule and hydrogen atoms are removed from the molecule.
Next, the molecule is converted into an InChI string, and the auxiliary information of the inchification
procedure is retrieved.
The N-layer is parsed and used to sort the atoms of the original order according
to the order in the InChI. In case, the molecule contains atoms that cannot be distinguished
with the InChI algorithm ('equivalent atoms'), the position of the unpaired electrons is changed
as to ensure the atoms with the lowest indices are used to compose the string.
"""
if mol.getRadicalCount() == 0 and not has_unexpected_lone_pairs(mol):
return None, None
elif mol.getFormula() == 'H':
return inchiutil.U_LAYER_PREFIX + '1', None
else:
molcopy = mol.copy(deep=True)
hydrogens = filter(lambda at: at.number == 1, molcopy.atoms)
[molcopy.removeAtom(h) for h in hydrogens]
rdkitmol = toRDKitMol(molcopy)
_, auxinfo = Chem.MolToInchiAndAuxInfo(rdkitmol, options='-SNon')# suppress stereo warnings
# extract the atom numbers from N-layer of auxiliary info:
atom_indices = inchiutil.parse_N_layer(auxinfo)
atom_indices = [atom_indices.index(i + 1) for i, atom in enumerate(molcopy.atoms)]
# sort the atoms based on the order of the atom indices
molcopy.atoms = [x for (y,x) in sorted(zip(atom_indices, molcopy.atoms), key=lambda pair: pair[0])]
ulayer = create_U_layer(molcopy, auxinfo)
player = create_P_layer(molcopy, auxinfo)
return ulayer, player