def test_rdkit_round_trip(self):
"""Test conversion to and from RDKitMol"""
for mol in self.test_mols:
rdkit_mol = toRDKitMol(mol)
new_mol = fromRDKitMol(Molecule(), rdkit_mol)
self.assertTrue(mol.isIsomorphic(new_mol))
self.assertEqual(mol.get_element_count(), new_mol.get_element_count())
def _rdkit_translator(input_object, identifier_type, mol=None):
"""
Converts between formats using RDKit. If input is a :class:`Molecule`,
the identifier_type is used to determine the output type. If the input is
a `str`, then the identifier_type is used to identify the input, and the
desired output is assumed to be a :class:`Molecule` object.
Args:
input_object: either molecule or string identifier
identifier_type: format of string identifier
'inchi' -> InChI
'inchikey' -> InChI Key
'sma' -> SMARTS
'smi' -> SMILES
mol: molecule object for output (optional)
"""
if identifier_type == 'inchi' and not Chem.inchi.INCHI_AVAILABLE:
raise DependencyError("RDKit installed without InChI. Please reinstall to read and write InChI strings.")
if isinstance(input_object, str):
# We are converting from a string identifier to a molecule
if identifier_type == 'inchi':
rdkitmol = Chem.inchi.MolFromInchi(input_object, removeHs=False)
elif identifier_type == 'sma':
rdkitmol = Chem.MolFromSmarts(input_object)
elif identifier_type == 'smi':
rdkitmol = Chem.MolFromSmiles(input_object)
else:
raise ValueError('Identifier type {0} is not supported for reading using RDKit.'.format(identifier_type))
if rdkitmol is None:
raise ValueError("Could not interpret the identifier {0!r}".format(input_object))
if mol is None:
mol = mm.Molecule()
output = fromRDKitMol(mol, rdkitmol)
elif isinstance(input_object, mm.Molecule):
# We are converting from a molecule to a string identifier
if identifier_type == 'smi':
rdkitmol = toRDKitMol(input_object, sanitize=False)
else:
rdkitmol = toRDKitMol(input_object, sanitize=True)
if identifier_type == 'inchi':
output = Chem.inchi.MolToInchi(rdkitmol, options='-SNon')
elif identifier_type == 'inchikey':
inchi = toInChI(input_object)
output = Chem.inchi.InchiToInchiKey(inchi)
elif identifier_type == 'sma':
output = Chem.MolToSmarts(rdkitmol)
elif identifier_type == 'smi':
if input_object.isAromatic():
output = Chem.MolToSmiles(rdkitmol)
else:
output = Chem.MolToSmiles(rdkitmol, kekuleSmiles=True)
else:
raise ValueError('Identifier type {0} is not supported for writing using RDKit.'.format(identifier_type))
else:
raise ValueError('Unexpected input format. Should be a Molecule or a string.')
return output
def _rdkit_translator(input_object, identifier_type, mol=None):
"""
Converts between formats using RDKit. If input is a :class:`Molecule`,
the identifier_type is used to determine the output type. If the input is
a `str`, then the identifier_type is used to identify the input, and the
desired output is assumed to be a :class:`Molecule` object.
Args:
input_object: either molecule or string identifier
identifier_type: format of string identifier
'inchi' -> InChI
'inchikey' -> InChI Key
'sma' -> SMARTS
'smi' -> SMILES
mol: molecule object for output (optional)
"""
if identifier_type == 'inchi' and not Chem.inchi.INCHI_AVAILABLE:
raise DependencyError("RDKit installed without InChI. Please reinstall to read and write InChI strings.")
if isinstance(input_object, str):
# We are converting from a string identifier to a molecule
if identifier_type == 'inchi':
rdkitmol = Chem.inchi.MolFromInchi(input_object, removeHs=False)
elif identifier_type == 'sma':
rdkitmol = Chem.MolFromSmarts(input_object)
elif identifier_type == 'smi':
rdkitmol = Chem.MolFromSmiles(input_object)
else:
raise ValueError('Identifier type {0} is not supported for reading using RDKit.'.format(identifier_type))
if rdkitmol is None:
raise ValueError("Could not interpret the identifier {0!r}".format(input_object))
if mol is None:
mol = mm.Molecule()
output = fromRDKitMol(mol, rdkitmol)
elif isinstance(input_object, mm.Molecule):
# We are converting from a molecule to a string identifier
if identifier_type == 'smi':
rdkitmol = toRDKitMol(input_object, sanitize=False)
else:
rdkitmol = toRDKitMol(input_object, sanitize=True)
if identifier_type == 'inchi':
output = Chem.inchi.MolToInchi(rdkitmol, options='-SNon')
elif identifier_type == 'inchikey':
inchi = toInChI(input_object)
output = Chem.inchi.InchiToInchiKey(inchi)
elif identifier_type == 'sma':
output = Chem.MolToSmarts(rdkitmol)
elif identifier_type == 'smi':
if input_object.isAromatic():
output = Chem.MolToSmiles(rdkitmol)
else:
output = Chem.MolToSmiles(rdkitmol, kekuleSmiles=True)
else:
raise ValueError('Identifier type {0} is not supported for writing using RDKit.'.format(identifier_type))
else:
raise ValueError('Unexpected input format. Should be a Molecule or a string.')
return output
def test_lone_pair_retention(self):
"""Test that we don't lose any lone pairs on round trip RDKit conversion."""
mol = Molecule().fromAdjacencyList("""
1 C u0 p0 c0 {2,D} {3,S} {4,S}
2 O u0 p2 c0 {1,D}
3 H u0 p0 c0 {1,S}
4 H u0 p0 c0 {1,S}
""")
rdmol = toRDKitMol(mol)
try:
mol2 = fromRDKitMol(Molecule(), rdmol)
except AtomTypeError as e:
self.fail('Could not convert from RDKitMol: ' + e.message)
else:
self.assertTrue(mol.isIsomorphic(mol2))
