def to_rdkit_mol(self):
"""
Return an :mod:`rdkit` representation.
Returns
-------
:class:`rdkit.Mol`
The molecule in :mod:`rdkit` format.
"""
mol = rdkit.EditableMol(rdkit.Mol())
for atom in self._atoms:
rdkit_atom = rdkit.Atom(atom.get_atomic_number())
rdkit_atom.SetFormalCharge(atom.get_charge())
mol.AddAtom(rdkit_atom)
for bond in self._bonds:
mol.AddBond(
beginAtomIdx=bond.get_atom1().get_id(),
endAtomIdx=bond.get_atom2().get_id(),
order=(rdkit.BondType.DATIVE if bond.get_order() == 9 else
rdkit.BondType(bond.get_order())),
)
mol = mol.GetMol()
rdkit_conf = rdkit.Conformer(len(self._atoms))
for atom_id, atom_coord in enumerate(self._position_matrix.T):
rdkit_conf.SetAtomPosition(atom_id, atom_coord)
mol.GetAtomWithIdx(atom_id).SetNoImplicit(True)
mol.AddConformer(rdkit_conf)
return mol
def to_rdkit_mol(self):
"""
Return an :mod:`rdkit` representation.
Returns
-------
:class:`rdkit.Mol`
The molecule in :mod:`rdkit` format.
"""
mol = rdkit.EditableMol(rdkit.Mol())
for atom in self.atoms:
rdkit_atom = rdkit.Atom(atom.atomic_number)
rdkit_atom.SetFormalCharge(atom.charge)
mol.AddAtom(rdkit_atom)
for bond in self.bonds:
mol.AddBond(beginAtomIdx=bond.atom1.id,
endAtomIdx=bond.atom2.id,
order=rdkit.BondType(bond.order))
mol = mol.GetMol()
rdkit_conf = rdkit.Conformer(len(self.atoms))
for atom_id, atom_coord in enumerate(self._position_matrix.T):
rdkit_conf.SetAtomPosition(atom_id, atom_coord)
mol.AddConformer(rdkit_conf)
return mol
def to_rdkit_mol_without_metals(mol, metal_atoms, metal_bonds):
"""
Create :class:`rdkit.Mol` with metals replaced by H atoms.
Parameters
----------
mol : :class:`.Molecule`
The molecule to be optimized.
metal_atoms : :class:`.list` of :class:`stk.Atom`
List of metal atoms.
metal_bonds : :class:`.list` of :class:`stk.Bond`
List of bonds including metal atoms.
Returns
-------
edit_mol : :class:`rdkit.Mol`
RDKit molecule with metal atoms replaced with H atoms.
"""
edit_mol = rdkit.EditableMol(rdkit.Mol())
for atom in mol.get_atoms():
if atom in metal_atoms:
# In place of metals, add H's that will be constrained.
# This allows the atom ids to not be changed.
rdkit_atom = rdkit.Atom(1)
rdkit_atom.SetFormalCharge(0)
else:
rdkit_atom = rdkit.Atom(atom.get_atomic_number())
rdkit_atom.SetFormalCharge(atom.get_charge())
edit_mol.AddAtom(rdkit_atom)
for bond in mol.get_bonds():
if bond in metal_bonds:
# Do not add bonds to metal atoms (replaced with H's).
continue
edit_mol.AddBond(
beginAtomIdx=bond.get_atom1().get_id(),
endAtomIdx=bond.get_atom2().get_id(),
order=rdkit.BondType(bond.get_order())
)
edit_mol = edit_mol.GetMol()
rdkit_conf = rdkit.Conformer(mol.get_num_atoms())
for atom_id, atom_coord in enumerate(mol.get_position_matrix()):
rdkit_conf.SetAtomPosition(atom_id, atom_coord)
edit_mol.GetAtomWithIdx(atom_id).SetNoImplicit(True)
edit_mol.AddConformer(rdkit_conf)
return edit_mol
def get_rdkit_mol(molecule):
rdkit_mol = rdkit.EditableMol(rdkit.Mol())
for atom in molecule.get_atoms():
rdkit_atom = rdkit.Atom(atom.get_atomic_number())
rdkit_atom.SetFormalCharge(atom.get_charge())
rdkit_mol.AddAtom(rdkit_atom)
for bond in molecule.get_bonds():
rdkit_mol.AddBond(
beginAtomIdx=bond.get_atom1().get_id(),
endAtomIdx=bond.get_atom2().get_id(),
order=(rdkit.BondType.DATIVE if bond.get_order() == 9 else
rdkit.BondType(bond.get_order())),
)
return rdkit_mol.GetMol()
def with_hydrogens(molecule):
rdkit_molecule = rdkit.EditableMol(rdkit.Mol())
for atom in molecule.get_atoms():
rdkit_atom = rdkit.Atom(atom.get_atomic_number())
rdkit_atom.SetFormalCharge(atom.get_charge())
rdkit_molecule.AddAtom(rdkit_atom)
for bond in molecule.get_bonds():
rdkit_molecule.AddBond(
beginAtomIdx=bond.get_atom1_id(),
endAtomIdx=bond.get_atom2_id(),
order=rdkit.BondType(bond.get_order()),
)
rdkit_molecule = rdkit_molecule.GetMol()
rdkit.SanitizeMol(rdkit_molecule)
return rdkit.AddHs(rdkit_molecule)
def vabene_to_smiles(molecule):
editable = rdkit.EditableMol(rdkit.Mol())
for atom in molecule.get_atoms():
rdkit_atom = rdkit.Atom(atom.get_atomic_number())
rdkit_atom.SetFormalCharge(atom.get_charge())
editable.AddAtom(rdkit_atom)
for bond in molecule.get_bonds():
editable.AddBond(
beginAtomIdx=bond.get_atom1_id(),
endAtomIdx=bond.get_atom2_id(),
order=rdkit.BondType(bond.get_order()),
)
rdkit_molecule = editable.GetMol()
rdkit.SanitizeMol(rdkit_molecule)
rdkit.Kekulize(rdkit_molecule)
return rdkit.MolToSmiles(rdkit_molecule)
def init_from_vabene_molecule(
cls,
molecule: vabene.Molecule,
functional_groups: typing.Iterable[typing.Union[
FunctionalGroup, FunctionalGroupFactory]] = (),
placer_ids: typing.Optional[tuple[int, ...]] = None,
position_matrix: typing.Optional[np.ndarray] = None,
) -> BuildingBlock:
"""
Initialize from a :mod:`vabene.Molecule`.
Notes:
The molecule is given 3D coordinates with
:func:`rdkit.ETKDGv2()`.
Parameters:
molecule:
The :class:`vabene.Molecule` from which to initialize.
functional_groups:
An :class:`iterable` of :class:`.FunctionalGroup` or
:class:`.FunctionalGroupFactory` or both.
:class:`.FunctionalGroup` instances are added to the
building block and :class:`.FunctionalGroupFactory`
instances are used to create :class:`.FunctionalGroup`
instances the building block should hold.
:class:`.FunctionalGroup` instances are used to
identify which atoms are modified during
:class:`.ConstructedMolecule` construction.
placer_ids:
The ids of *placer* atoms. These are the atoms which
should be used for calculating the position of the
building block. Depending on the values passed to
`placer_ids`, and the functional groups in the building
block, different *placer* ids will be used by the
building block.
#. `placer_ids` is passed to the initializer: the
passed *placer* ids will be used by the building
block.
#. `placer_ids` is ``None`` and the building block has
functional groups: The *placer* ids of the
functional groups will be used as the *placer* ids
of the building block.
#. `placer_ids` is ``None`` and `functional_groups` is
empty. All atoms of the molecule will be used for
*placer* ids.
position_matrix:
The position matrix the building block should use. If
``None``, :func:`rdkit.ETKDGv2` will be used to
calculate it.
Returns:
The building block.
Raises:
:class:`RuntimeError`
If embedding the molecule fails.
"""
editable = rdkit.EditableMol(rdkit.Mol())
for atom in molecule.get_atoms():
rdkit_atom = rdkit.Atom(atom.get_atomic_number())
rdkit_atom.SetFormalCharge(atom.get_charge())
editable.AddAtom(rdkit_atom)
for bond in molecule.get_bonds():
editable.AddBond(
beginAtomIdx=bond.get_atom1_id(),
endAtomIdx=bond.get_atom2_id(),
order=rdkit.BondType(bond.get_order()),
)
rdkit_molecule = editable.GetMol()
rdkit.SanitizeMol(rdkit_molecule)
rdkit_molecule = rdkit.AddHs(rdkit_molecule)
if position_matrix is None:
params = rdkit.ETKDGv2()
random_seed = 4
params.randomSeed = random_seed
if rdkit.EmbedMolecule(rdkit_molecule, params) == -1:
raise RuntimeError(
f'Embedding with seed value of {random_seed} '
'failed.')
else:
# Make sure the position matrix always holds floats.
position_matrix = np.array(
position_matrix,
dtype=np.float64,
)
conformer = rdkit.Conformer(rdkit_molecule.GetNumAtoms())
for atom_id, position in enumerate(position_matrix):
conformer.SetAtomPosition(atom_id, position)
rdkit_molecule.AddConformer(conformer)
rdkit.Kekulize(rdkit_molecule)
return cls.init_from_rdkit_mol(
molecule=rdkit_molecule,
functional_groups=functional_groups,
placer_ids=placer_ids,
)
def init_from_vabene_molecule(
cls,
molecule,
functional_groups=(),
placer_ids=None,
random_seed=None,
position_matrix=None,
):
"""
Initialize from a :mod:`vabene.Molecule`.
Notes
-----
The molecule is given 3D coordinates with
:func:`rdkit.ETKDGv2()`.
Parameters
----------
molecule : :class:`vabene.Molecule`
The molecule from which to initialize.
functional_groups : :class:`iterable`, optional
An :class:`iterable` of :class:`.FunctionalGroup` or
:class:`.FunctionalGroupFactory` or both.
:class:`.FunctionalGroup` instances are added to the
building block and :class:`.FunctionalGroupFactory`
instances are used to create :class:`.FunctionalGroup`
instances the building block should hold.
:class:`.FunctionalGroup` instances are used to identify
which atoms are modified during
:class:`.ConstructedMolecule` construction.
placer_ids : :class:`tuple` of :class:`int`, optional
The ids of *placer* atoms. These are the atoms which should
be used for calculating the position of the building block.
Depending on the values passed to `placer_ids`,
and the functional groups in the building block, different
*placer* ids will be used by the building block.
#. `placer_ids` is passed to the initializer: the passed
*placer* ids will be used by the building block.
#. `placer_ids` is ``None`` and the building block has
functional groups: The *placer* ids of the functional
groups will be used as the *placer* ids of the building
block.
#. `placer_ids` is ``None`` and `functional_groups` is
empty. All atoms of the molecule will be used for
*placer* ids.
random_seed : :class:`int`, optional
Random seed passed to :func:`rdkit.ETKDGv2`. This
argument is deprecated and will be removed in any
version of :mod:`stk` released on, or after, 01/08/20.
If you want to change the position matrix of the
initialized building block, please use
`position_matrix`.
position_matrix : :class:`numpy.ndarray`, optional
The position matrix the building block should use. If
``None``, :func:`rdkit.ETKDGv2` will be used to calculate
it.
Returns
-------
:class:`.BuildingBlock`
The building block.
Raises
------
:class:`RuntimeError`
If embedding the molecule fails.
"""
if random_seed is not None:
warnings.warn(
'The init_from_vabene_molecule random seed parameter '
'will be removed in any '
'version of stk released on, or after, 01/08/20. '
'If you want to change the position matrix of the '
'initialized building block, please use the '
'position_matrix parameter.',
FutureWarning,
)
else:
random_seed = 4
editable = rdkit.EditableMol(rdkit.Mol())
for atom in molecule.get_atoms():
rdkit_atom = rdkit.Atom(atom.get_atomic_number())
rdkit_atom.SetFormalCharge(atom.get_charge())
editable.AddAtom(rdkit_atom)
for bond in molecule.get_bonds():
editable.AddBond(
beginAtomIdx=bond.get_atom1_id(),
endAtomIdx=bond.get_atom2_id(),
order=rdkit.BondType(bond.get_order()),
)
rdkit_molecule = editable.GetMol()
rdkit.SanitizeMol(rdkit_molecule)
rdkit_molecule = rdkit.AddHs(rdkit_molecule)
if position_matrix is None:
params = rdkit.ETKDGv2()
params.randomSeed = random_seed
if rdkit.EmbedMolecule(rdkit_molecule, params) == -1:
raise RuntimeError(
f'Embedding with seed value of {random_seed} '
'failed.'
)
else:
# Make sure the position matrix always holds floats.
position_matrix = np.array(
position_matrix,
dtype=np.float64,
)
conformer = rdkit.Conformer(rdkit_molecule.GetNumAtoms())
for atom_id, position in enumerate(position_matrix):
conformer.SetAtomPosition(atom_id, position)
rdkit_molecule.AddConformer(conformer)
rdkit.Kekulize(rdkit_molecule)
return cls.init_from_rdkit_mol(
molecule=rdkit_molecule,
functional_groups=functional_groups,
placer_ids=placer_ids,
)
