def generate_conformers(rdkit_mol: Chem.Mol,
conformer_no: int) -> List[np.ndarray]:
"""
Generate a set of conformers for the molecule including the input conformer.
Args:
rdkit_mol:
A complete Chem.Mol instance of a molecule.
conformer_no:
The number of conformers made for the molecule
return:
A list of conformer position arrays
"""
AllChem.EmbedMultipleConfs(
rdkit_mol,
numConfs=conformer_no,
randomSeed=1,
clearConfs=False,
useBasicKnowledge=True,
pruneRmsThresh=1,
enforceChirality=True,
)
positions = rdkit_mol.GetConformers()
return [conformer.GetPositions() for conformer in positions]
def prune_conformers(
mol: Chem.Mol, energies: list, rmsd_threshold: float
) -> (Chem.Mol, list):
"""
Adopted from: https://github.com/skearnes/rdkit-utils/blob/master/rdkit_utils/conformers.py
Prune conformers from a molecule using an RMSD threshold, starting
with the lowest energy conformer.
Parameters
----------
mol : RDKit Mol
Molecule.
Returns
-------
A new RDKit Mol containing the chosen conformers, sorted by
increasing energy.
"""
from typing import List
rmsd = get_conformer_rmsd(mol)
sort = np.argsort(
[x.value_in_unit(unit.kilocalorie_per_mole) for x in energies]
) # sort by increasing energy
keep: List[int] = [] # always keep lowest-energy conformer
discard: List[int] = []
for i in sort:
# always keep lowest-energy conformer
if len(keep) == 0:
keep.append(i)
continue
# get RMSD to selected conformers
this_rmsd = rmsd[i][np.asarray(keep, dtype=int)]
# discard conformers within the RMSD threshold
if np.all(this_rmsd >= rmsd_threshold):
keep.append(i)
else:
discard.append(i)
# create a new molecule to hold the chosen conformers
# this ensures proper conformer IDs and energy-based ordering
new_mol = Chem.Mol(mol)
new_mol.RemoveAllConformers()
conf_ids = [conf.GetId() for conf in mol.GetConformers()]
filtered_energies = []
logger.debug(f"keep: {keep}")
for i in keep:
logger.debug(i)
conf = mol.GetConformer(conf_ids[i])
filtered_energies.append(energies[i])
new_mol.AddConformer(conf, assignId=True)
return new_mol, filtered_energies
def generate_conformers(rdkit_mol: Chem.Mol,
conformer_no=10) -> List[np.ndarray]:
"""
Generate a set of x conformers of the molecule
:param conformer_no: The amount of conformers made for the molecule
:param rdkit_mol: The name of the input file
:return: A list of conformer position arrays
"""
AllChem.EmbedMultipleConfs(rdkit_mol, numConfs=conformer_no)
positions = rdkit_mol.GetConformers()
return [conformer.GetPositions() for conformer in positions]
def draw(mol: Chem.Mol, energies: list = []) -> interact:
"""Make a drawing of all conformers in 3d
:param mol: rdkit molecule
:type mol: rdkit.Chem.Mol
:param energies: list of conformer energies
:type energies: list
:return: interact 3D object
"""
p = py3Dmol.view(width=400, height=400)
return interact(_graph_conf,
m=fixed(mol),
p=fixed(p),
confId=[c.GetId() for c in mol.GetConformers()],
energies=fixed(energies))
def extract_from_rdmol(mol: Chem.Mol) -> tuple:
"""Extract information from RDKit Mol object with conformers.
:param mol: rdkit molecule
:type mol: rdkit.Chem.Mol
:return: tuple(elements, conformer_coordinates, connectivity_matrix, charges)
"""
# Take out elements, coordinates and connectivity matrix
elements = [atom.GetSymbol() for atom in mol.GetAtoms()]
charges = np.array([atom.GetFormalCharge() for atom in mol.GetAtoms()])
conformer_coordinates = []
for conformer in mol.GetConformers():
coordinates = conformer.GetPositions()
conformer_coordinates.append(coordinates)
conformer_coordinates = np.array(conformer_coordinates)
connectivity_matrix = Chem.GetAdjacencyMatrix(mol, useBO=True)
return elements, conformer_coordinates, connectivity_matrix, charges
def cluster_by_rmsd(m: Chem.Mol, energies, rmsd_threshold: float):
"""
conformer in m should be assigned consecutive ids with delta=1
:param energies:
:param m:
:param rmsd_threshold:
:return:
"""
rmsd_condensed = AllChem.GetConformerRMSMatrix(m)
# rmsd_condensed = ConfGen.GetBestRMSMatrix(m)
from rdkit.ML.Cluster.Butina import ClusterData
clusters = ClusterData(rmsd_condensed,
len(m.GetConformers()),
distThresh=rmsd_threshold,
isDistData=True)
m_after_prune = Chem.Mol(m)
m_after_prune.RemoveAllConformers()
n_energies = []
for c in clusters:
conf = m.GetConformer(c[0])
n_energies.append(energies[c[0]])
m_after_prune.AddConformer(conf, assignId=True)
return m_after_prune, n_energies