def get_canonical_smiles_from_mol2(mol2_dict):
canonical_smiles_dict = {}
for i, (mol_name, mol2_file) in enumerate(mol2_dict.items()):
mol = mol_from_mol2(mol2_file, mol_name)
try:
mol = mol_to_standardised_mol(mol)
except:
pass
canon_smiles = MolToSmiles(mol, isomericSmiles=True)
canonical_smiles_dict[mol_name] = canon_smiles
return canonical_smiles_dict
def test_standardisation(self):
import rdkit.Chem
from e3fp.conformer.util import (
mol_from_smiles,
mol_to_standardised_mol,
)
smiles = "C[N-]c1cccc[n+]1C"
mol = mol_from_smiles(smiles, "tmp")
self.assertEqual(rdkit.Chem.MolToSmiles(mol), smiles)
mol = mol_to_standardised_mol(mol)
self.assertEqual(rdkit.Chem.MolToSmiles(mol), "CN=c1ccccn1C")
def generate_conformers(
input_mol,
name=None,
standardise=STANDARDISE_DEF,
num_conf=NUM_CONF_DEF,
first=FIRST_DEF,
pool_multiplier=POOL_MULTIPLIER_DEF,
rmsd_cutoff=RMSD_CUTOFF_DEF,
max_energy_diff=MAX_ENERGY_DIFF_DEF,
forcefield=FORCEFIELD_DEF,
seed=SEED_DEF,
out_file=None,
out_dir=OUTDIR_DEF,
save=False,
compress=COMPRESS_DEF,
overwrite=False,
):
"""Generate and save conformers for molecules.
Parameters
----------
input_mol : RDKit Mol
Mol with a single conformer from which to generate conformers.
name : str, optional
Name of molecule.
standardise : bool, optional
Standardise mol before generating conformers.
num_conf : int, optional
If int, this is the target number of conformations. If -1, number
of conformations is automatically chosen based on number of rotatable
bonds.
first : int, optional
Number of first conformers to return. Does not impact conformer
generator process, except may terminate conformer generation early when
this many of conformers have been accepted.
pool_multiplier : int, optional
Factor to multiply by `num_conf`. The resulting number of conformations
will be generated, then pruned to `num_conf`.
rmsd_cutoff : float, optional
RMSD threshold above which to accept two conformations as different
max_energy_diff : float, optional
Maximum energy difference between lowest energy conformer and any
accepted conformer.
forcefield : {'uff', 'mmff94', 'mmff94s'}, optional
Forcefield to use for minimization of conformers.
seed : int, optional
Random seed for conformer generation. If -1, the random number
generator is unseeded.
out_file : str, optional
Filename to save output, if `save` is True. If None, filename will be
`name`.sdf, optionally with a compressed extension.
out_dir : str, optional
Directory where output files will be saved if `save` is True.
save : bool, optional
Save conformers to `out_file` in `out_dir`.
compress : int, optional
Compression of SDF files.
None: auto. Mode is chosen based on extension, defaulting to SDF.
0: File is not compressed.
1: File is gzipped (.gz)
2: File is bzipped (.bz2)
overwrite : bool, optional
Overwrite output files if they already exist.
Returns
-------
bool
If something went wrong, only return False. Otherwise return below.
tuple
Tuple with molecule name, number of rotatable bonds, numpy array of
indices of final conformations, numpy array of energies of all
conformations generated, and 2D numpy array of pairwise RMSDs between
final conformations.
"""
if name is None:
name = input_mol.GetProp("_Name")
if standardise:
input_mol = mol_to_standardised_mol(input_mol)
if save:
if out_file is None:
extensions = ("", ".gz", ".bz2")
if compress not in (0, 1, 2):
compress = 0
out_file = os.path.join(
out_dir, "{}.sdf{}".format(name, extensions[compress]))
if os.path.exists(out_file) and not overwrite:
logging.warning("{} already exists. Skipping.".format(out_file))
return False
logging.info("Generating conformers for {}.".format(name))
try:
conf_gen = ConformerGenerator(
num_conf=num_conf,
first=first,
pool_multiplier=pool_multiplier,
rmsd_cutoff=rmsd_cutoff,
max_energy_diff=max_energy_diff,
forcefield=forcefield,
seed=seed,
get_values=True,
)
mol, values = conf_gen.generate_conformers(input_mol)
logging.info("Generated {:d} conformers for {}.".format(
mol.GetNumConformers(), name))
except Exception:
logging.warning("Problem generating conformers for {}.".format(name),
exc_info=True)
return False
if save:
try:
mol_to_sdf(mol, out_file)
logging.info("Saved conformers for {} to {}.".format(
name, out_file))
except Exception:
logging.warning(
"Problem saving conformers for {} to {}.".format(
name, out_file),
exc_info=True,
)
return (mol, name, AllChem.CalcNumRotatableBonds(mol)) + values