Loading deepchem/dock/pose_generation.py +5 −4 Original line number Diff line number Diff line Loading @@ -16,7 +16,7 @@ from deepchem.utils.data_utils import download_url, get_data_dir from deepchem.utils.typing import RDKitMol from deepchem.utils.geometry_utils import compute_centroid, compute_protein_range from deepchem.utils.rdkit_utils import load_molecule, write_molecule from deepchem.utils.vina_utils import load_docked_ligands, write_vina_conf, write_gnina_conf, read_gnina_log from deepchem.utils.docking_utils import load_docked_ligands, write_vina_conf, write_gnina_conf, read_gnina_log logger = logging.getLogger(__name__) DOCKED_POSES = List[Tuple[RDKitMol, RDKitMol]] Loading Loading @@ -102,8 +102,9 @@ class GninaPoseGenerator(PoseGenerator): "Protein–Ligand Scoring with Convolutional Neural Networks." Journal of chemical information and modeling (2017). Notes ----- Note ---- * GNINA currently only works on Linux operating systems. * GNINA requires CUDA >= 10.1 for fast CNN scoring. * Almost all dependencies are included in the most compatible way possible, which reduces performance. Build GNINA from source Loading @@ -122,7 +123,7 @@ class GninaPoseGenerator(PoseGenerator): self.gnina_cmd = os.path.join(self.gnina_dir, filename) else: raise ValueError( "Unknown operating system. Try using a cloud platform to run this code instead." "GNINA currently only runs on Linux. Try using a cloud platform to run this code instead." ) if not os.path.exists(self.gnina_cmd): Loading deepchem/dock/tests/test_pose_generation.py +3 −1 Original line number Diff line number Diff line Loading @@ -11,6 +11,7 @@ import deepchem as dc import pytest IS_WINDOWS = platform.system() == 'Windows' IS_LINUX = platform.system() == 'Linux' class TestPoseGeneration(unittest.TestCase): Loading @@ -23,7 +24,7 @@ class TestPoseGeneration(unittest.TestCase): """Test that VinaPoseGenerator can be initialized.""" dc.dock.VinaPoseGenerator() @unittest.skipIf(IS_WINDOWS, 'Skip the test on Windows') @unittest.skipIf(not IS_LINUX, 'Skip the test on Windows and Mac.') def test_gnina_initialization(self): """Test that GninaPoseGenerator can be initialized.""" dc.dock.GninaPoseGenerator() Loading Loading @@ -64,6 +65,7 @@ class TestPoseGeneration(unittest.TestCase): assert isinstance(ligand, Chem.Mol) @pytest.mark.slow @unittest.skipIf(not IS_LINUX, 'Skip the test on Windows and Mac.') def test_gnina_poses_and_scores(self): """Test that GninaPoseGenerator generates poses and scores Loading deepchem/utils/__init__.py +5 −5 Original line number Diff line number Diff line Loading @@ -83,11 +83,11 @@ from deepchem.utils.pdbqt_utils import pdbqt_to_pdb from deepchem.utils.pdbqt_utils import convert_protein_to_pdbqt from deepchem.utils.pdbqt_utils import convert_mol_to_pdbqt from deepchem.utils.vina_utils import write_vina_conf from deepchem.utils.vina_utils import write_gnina_conf from deepchem.utils.vina_utils import read_gnina_log from deepchem.utils.vina_utils import load_docked_ligands from deepchem.utils.vina_utils import prepare_inputs from deepchem.utils.docking_utils import write_vina_conf from deepchem.utils.docking_utils import write_gnina_conf from deepchem.utils.docking_utils import read_gnina_log from deepchem.utils.docking_utils import load_docked_ligands from deepchem.utils.docking_utils import prepare_inputs from deepchem.utils.voxel_utils import convert_atom_to_voxel from deepchem.utils.voxel_utils import convert_atom_pair_to_voxel Loading deepchem/utils/docking_utils.py 0 → 100644 +308 −0 Original line number Diff line number Diff line """ This file contains utilities for molecular docking. """ from typing import List, Optional, Tuple import os import numpy as np from deepchem.utils.typing import RDKitMol from deepchem.utils.pdbqt_utils import pdbqt_to_pdb def write_vina_conf(protein_filename: str, ligand_filename: str, centroid: np.ndarray, box_dims: np.ndarray, conf_filename: str, num_modes: int = 9, exhaustiveness: int = None) -> None: """Writes Vina configuration file to disk. Autodock Vina accepts a configuration file which provides options under which Vina is invoked. This utility function writes a vina configuration file which directs Autodock vina to perform docking under the provided options. Parameters ---------- protein_filename: str Filename for protein ligand_filename: str Filename for the ligand centroid: np.ndarray A numpy array with shape `(3,)` holding centroid of system box_dims: np.ndarray A numpy array of shape `(3,)` holding the size of the box to dock conf_filename: str Filename to write Autodock Vina configuration to. num_modes: int, optional (default 9) The number of binding modes Autodock Vina should find exhaustiveness: int, optional The exhaustiveness of the search to be performed by Vina """ with open(conf_filename, "w") as f: f.write("receptor = %s\n" % protein_filename) f.write("ligand = %s\n\n" % ligand_filename) f.write("center_x = %f\n" % centroid[0]) f.write("center_y = %f\n" % centroid[1]) f.write("center_z = %f\n\n" % centroid[2]) f.write("size_x = %f\n" % box_dims[0]) f.write("size_y = %f\n" % box_dims[1]) f.write("size_z = %f\n\n" % box_dims[2]) f.write("num_modes = %d\n\n" % num_modes) if exhaustiveness is not None: f.write("exhaustiveness = %d\n" % exhaustiveness) def write_gnina_conf(protein_filename: str, ligand_filename: str, conf_filename: str, num_modes: int = 9, exhaustiveness: int = None, **kwargs) -> None: """Writes GNINA configuration file to disk. GNINA accepts a configuration file which provides options under which GNINA is invoked. This utility function writes a configuration file which directs GNINA to perform docking under the provided options. Parameters ---------- protein_filename: str Filename for protein ligand_filename: str Filename for the ligand conf_filename: str Filename to write Autodock Vina configuration to. num_modes: int, optional (default 9) The number of binding modes GNINA should find exhaustiveness: int, optional The exhaustiveness of the search to be performed by GNINA kwargs: Args supported by GNINA documented here https://github.com/gnina/gnina#usage """ with open(conf_filename, "w") as f: f.write("receptor = %s\n" % protein_filename) f.write("ligand = %s\n\n" % ligand_filename) f.write("autobox_ligand = %s\n\n" % protein_filename) if exhaustiveness is not None: f.write("exhaustiveness = %d\n" % exhaustiveness) f.write("num_modes = %d\n\n" % num_modes) for k, v in kwargs.items(): f.write("%s = %s\n" % (str(k), str(v))) def read_gnina_log(log_file: str) -> np.array: """Read GNINA logfile and get docking scores. GNINA writes computed binding affinities to a logfile. Parameters ---------- log_file: str Filename of logfile generated by GNINA. Returns ------- scores: np.array, dimension (num_modes, 3) Array of binding affinity (kcal/mol), CNN pose score, and CNN affinity for each binding mode. """ scores = [] lines = open(log_file).readlines() mode_start = np.inf for idx, line in enumerate(lines): if line[:6] == '-----+': mode_start = idx if idx > mode_start: mode = line.split() score = [float(x) for x in mode[1:]] scores.append(score) scores = np.array(scores) return scores def load_docked_ligands( pdbqt_output: str) -> Tuple[List[RDKitMol], List[float]]: """This function loads ligands docked by autodock vina. Autodock vina writes outputs to disk in a PDBQT file format. This PDBQT file can contain multiple docked "poses". Recall that a pose is an energetically favorable 3D conformation of a molecule. This utility function reads and loads the structures for multiple poses from vina's output file. Parameters ---------- pdbqt_output: str Should be the filename of a file generated by autodock vina's docking software. Returns ------- Tuple[List[rdkit.Chem.rdchem.Mol], List[float]] Tuple of `molecules, scores`. `molecules` is a list of rdkit molecules with 3D information. `scores` is the associated vina score. Notes ----- This function requires RDKit to be installed. """ try: from rdkit import Chem except ModuleNotFoundError: raise ImportError("This function requires RDKit to be installed.") lines = open(pdbqt_output).readlines() molecule_pdbqts = [] scores = [] current_pdbqt: Optional[List[str]] = None for line in lines: if line[:5] == "MODEL": current_pdbqt = [] elif line[:19] == "REMARK VINA RESULT:": words = line.split() # the line has format # REMARK VINA RESULT: score ... # There is only 1 such line per model so we can append it scores.append(float(words[3])) elif line[:6] == "ENDMDL": molecule_pdbqts.append(current_pdbqt) current_pdbqt = None else: # FIXME: Item "None" of "Optional[List[str]]" has no attribute "append" current_pdbqt.append(line) # type: ignore molecules = [] for pdbqt_data in molecule_pdbqts: pdb_block = pdbqt_to_pdb(pdbqt_data=pdbqt_data) mol = Chem.MolFromPDBBlock(str(pdb_block), sanitize=False, removeHs=False) molecules.append(mol) return molecules, scores def prepare_inputs(protein: str, ligand: str, replace_nonstandard_residues: bool = True, remove_heterogens: bool = True, remove_water: bool = True, add_hydrogens: bool = True, pH: float = 7.0, optimize_ligand: bool = True, pdb_name: Optional[str] = None) -> Tuple[RDKitMol, RDKitMol]: """This prepares protein-ligand complexes for docking. Autodock Vina requires PDB files for proteins and ligands with sensible inputs. This function uses PDBFixer and RDKit to ensure that inputs are reasonable and ready for docking. Default values are given for convenience, but fixing PDB files is complicated and human judgement is required to produce protein structures suitable for docking. Always inspect the results carefully before trying to perform docking. Parameters ---------- protein: str Filename for protein PDB file or a PDBID. ligand: str Either a filename for a ligand PDB file or a SMILES string. replace_nonstandard_residues: bool (default True) Replace nonstandard residues with standard residues. remove_heterogens: bool (default True) Removes residues that are not standard amino acids or nucleotides. remove_water: bool (default True) Remove water molecules. add_hydrogens: bool (default True) Add missing hydrogens at the protonation state given by `pH`. pH: float (default 7.0) Most common form of each residue at given `pH` value is used. optimize_ligand: bool (default True) If True, optimize ligand with RDKit. Required for SMILES inputs. pdb_name: Optional[str] If given, write sanitized protein and ligand to files called "pdb_name.pdb" and "ligand_pdb_name.pdb" Returns ------- Tuple[RDKitMol, RDKitMol] Tuple of `protein_molecule, ligand_molecule` with 3D information. Note ---- This function requires RDKit and OpenMM to be installed. Read more about PDBFixer here: https://github.com/openmm/pdbfixer. Examples -------- >>> p, m = prepare_inputs('3cyx', 'CCC') >>> p.GetNumAtoms() 1415 >>> m.GetNumAtoms() 11 >>> p, m = prepare_inputs('3cyx', 'CCC', remove_heterogens=False) >>> p.GetNumAtoms() 1720 """ try: from rdkit import Chem from rdkit.Chem import AllChem from pdbfixer import PDBFixer from simtk.openmm.app import PDBFile except ModuleNotFoundError: raise ImportError( "This function requires RDKit and OpenMM to be installed.") if protein.endswith('.pdb'): fixer = PDBFixer(protein) else: fixer = PDBFixer(url='https://files.rcsb.org/download/%s.pdb' % (protein)) if ligand.endswith('.pdb'): m = Chem.MolFromPDBFile(ligand) else: m = Chem.MolFromSmiles(ligand, sanitize=True) # Apply common fixes to PDB files if replace_nonstandard_residues: fixer.findMissingResidues() fixer.findNonstandardResidues() fixer.replaceNonstandardResidues() if remove_heterogens and not remove_water: fixer.removeHeterogens(True) if remove_heterogens and remove_water: fixer.removeHeterogens(False) if add_hydrogens: fixer.addMissingHydrogens(pH) PDBFile.writeFile(fixer.topology, fixer.positions, open('tmp.pdb', 'w')) p = Chem.MolFromPDBFile('tmp.pdb', sanitize=True) os.remove('tmp.pdb') # Optimize ligand if optimize_ligand: m = Chem.AddHs(m) # need hydrogens for optimization AllChem.EmbedMolecule(m) AllChem.MMFFOptimizeMolecule(m) if pdb_name: Chem.rdmolfiles.MolToPDBFile(p, '%s.pdb' % (pdb_name)) Chem.rdmolfiles.MolToPDBFile(m, 'ligand_%s.pdb' % (pdb_name)) return (p, m) deepchem/utils/test/test_vina_utils.py→deepchem/utils/test/test_docking_utils.py +8 −8 Original line number Diff line number Diff line Loading @@ -4,7 +4,7 @@ Test Autodock Vina Utility Functions. import os import numpy as np import unittest from deepchem.utils import vina_utils from deepchem.utils import docking_utils from deepchem.utils import rdkit_utils Loading @@ -17,7 +17,8 @@ class TestVinaUtils(unittest.TestCase): '1jld_ligand_docked.pdbqt') def test_load_docked_ligand(self): docked_ligands, scores = vina_utils.load_docked_ligands(self.docked_ligands) docked_ligands, scores = docking_utils.load_docked_ligands( self.docked_ligands) assert len(docked_ligands) == 9 assert len(scores) == 9 Loading @@ -27,7 +28,7 @@ class TestVinaUtils(unittest.TestCase): assert np.count_nonzero(xyz) > 0 def test_write_gnina_conf(self): vina_utils.write_gnina_conf( docking_utils.write_gnina_conf( 'protein.pdb', 'ligand.sdf', 'conf.txt', Loading @@ -37,7 +38,7 @@ class TestVinaUtils(unittest.TestCase): def test_read_gnina_log(self): log_file = os.path.join(self.current_dir, 'data', 'gnina_log.txt') scores = vina_utils.read_gnina_log(log_file) scores = docking_utils.read_gnina_log(log_file) assert np.array_equal( scores, np.array([[-4.37, 0.6392, 4.336], [-3.56, 0.6202, 4.162]])) Loading @@ -45,13 +46,13 @@ class TestVinaUtils(unittest.TestCase): pdbid = '3cyx' ligand_smiles = 'CC(C)(C)NC(O)C1CC2CCCCC2C[NH+]1CC(O)C(CC1CCCCC1)NC(O)C(CC(N)O)NC(O)C1CCC2CCCCC2N1' protein, ligand = vina_utils.prepare_inputs( protein, ligand = docking_utils.prepare_inputs( pdbid, ligand_smiles, pdb_name=pdbid) assert np.isclose(protein.GetNumAtoms(), 1415, atol=3) assert np.isclose(ligand.GetNumAtoms(), 124, atol=3) protein, ligand = vina_utils.prepare_inputs(pdbid + '.pdb', protein, ligand = docking_utils.prepare_inputs(pdbid + '.pdb', 'ligand_' + pdbid + '.pdb') assert np.isclose(protein.GetNumAtoms(), 1415, atol=3) Loading @@ -59,4 +60,3 @@ class TestVinaUtils(unittest.TestCase): os.remove(pdbid + '.pdb') os.remove('ligand_' + pdbid + '.pdb') os.remove('tmp.pdb') Loading
deepchem/dock/pose_generation.py +5 −4 Original line number Diff line number Diff line Loading @@ -16,7 +16,7 @@ from deepchem.utils.data_utils import download_url, get_data_dir from deepchem.utils.typing import RDKitMol from deepchem.utils.geometry_utils import compute_centroid, compute_protein_range from deepchem.utils.rdkit_utils import load_molecule, write_molecule from deepchem.utils.vina_utils import load_docked_ligands, write_vina_conf, write_gnina_conf, read_gnina_log from deepchem.utils.docking_utils import load_docked_ligands, write_vina_conf, write_gnina_conf, read_gnina_log logger = logging.getLogger(__name__) DOCKED_POSES = List[Tuple[RDKitMol, RDKitMol]] Loading Loading @@ -102,8 +102,9 @@ class GninaPoseGenerator(PoseGenerator): "Protein–Ligand Scoring with Convolutional Neural Networks." Journal of chemical information and modeling (2017). Notes ----- Note ---- * GNINA currently only works on Linux operating systems. * GNINA requires CUDA >= 10.1 for fast CNN scoring. * Almost all dependencies are included in the most compatible way possible, which reduces performance. Build GNINA from source Loading @@ -122,7 +123,7 @@ class GninaPoseGenerator(PoseGenerator): self.gnina_cmd = os.path.join(self.gnina_dir, filename) else: raise ValueError( "Unknown operating system. Try using a cloud platform to run this code instead." "GNINA currently only runs on Linux. Try using a cloud platform to run this code instead." ) if not os.path.exists(self.gnina_cmd): Loading
deepchem/dock/tests/test_pose_generation.py +3 −1 Original line number Diff line number Diff line Loading @@ -11,6 +11,7 @@ import deepchem as dc import pytest IS_WINDOWS = platform.system() == 'Windows' IS_LINUX = platform.system() == 'Linux' class TestPoseGeneration(unittest.TestCase): Loading @@ -23,7 +24,7 @@ class TestPoseGeneration(unittest.TestCase): """Test that VinaPoseGenerator can be initialized.""" dc.dock.VinaPoseGenerator() @unittest.skipIf(IS_WINDOWS, 'Skip the test on Windows') @unittest.skipIf(not IS_LINUX, 'Skip the test on Windows and Mac.') def test_gnina_initialization(self): """Test that GninaPoseGenerator can be initialized.""" dc.dock.GninaPoseGenerator() Loading Loading @@ -64,6 +65,7 @@ class TestPoseGeneration(unittest.TestCase): assert isinstance(ligand, Chem.Mol) @pytest.mark.slow @unittest.skipIf(not IS_LINUX, 'Skip the test on Windows and Mac.') def test_gnina_poses_and_scores(self): """Test that GninaPoseGenerator generates poses and scores Loading
deepchem/utils/__init__.py +5 −5 Original line number Diff line number Diff line Loading @@ -83,11 +83,11 @@ from deepchem.utils.pdbqt_utils import pdbqt_to_pdb from deepchem.utils.pdbqt_utils import convert_protein_to_pdbqt from deepchem.utils.pdbqt_utils import convert_mol_to_pdbqt from deepchem.utils.vina_utils import write_vina_conf from deepchem.utils.vina_utils import write_gnina_conf from deepchem.utils.vina_utils import read_gnina_log from deepchem.utils.vina_utils import load_docked_ligands from deepchem.utils.vina_utils import prepare_inputs from deepchem.utils.docking_utils import write_vina_conf from deepchem.utils.docking_utils import write_gnina_conf from deepchem.utils.docking_utils import read_gnina_log from deepchem.utils.docking_utils import load_docked_ligands from deepchem.utils.docking_utils import prepare_inputs from deepchem.utils.voxel_utils import convert_atom_to_voxel from deepchem.utils.voxel_utils import convert_atom_pair_to_voxel Loading
deepchem/utils/docking_utils.py 0 → 100644 +308 −0 Original line number Diff line number Diff line """ This file contains utilities for molecular docking. """ from typing import List, Optional, Tuple import os import numpy as np from deepchem.utils.typing import RDKitMol from deepchem.utils.pdbqt_utils import pdbqt_to_pdb def write_vina_conf(protein_filename: str, ligand_filename: str, centroid: np.ndarray, box_dims: np.ndarray, conf_filename: str, num_modes: int = 9, exhaustiveness: int = None) -> None: """Writes Vina configuration file to disk. Autodock Vina accepts a configuration file which provides options under which Vina is invoked. This utility function writes a vina configuration file which directs Autodock vina to perform docking under the provided options. Parameters ---------- protein_filename: str Filename for protein ligand_filename: str Filename for the ligand centroid: np.ndarray A numpy array with shape `(3,)` holding centroid of system box_dims: np.ndarray A numpy array of shape `(3,)` holding the size of the box to dock conf_filename: str Filename to write Autodock Vina configuration to. num_modes: int, optional (default 9) The number of binding modes Autodock Vina should find exhaustiveness: int, optional The exhaustiveness of the search to be performed by Vina """ with open(conf_filename, "w") as f: f.write("receptor = %s\n" % protein_filename) f.write("ligand = %s\n\n" % ligand_filename) f.write("center_x = %f\n" % centroid[0]) f.write("center_y = %f\n" % centroid[1]) f.write("center_z = %f\n\n" % centroid[2]) f.write("size_x = %f\n" % box_dims[0]) f.write("size_y = %f\n" % box_dims[1]) f.write("size_z = %f\n\n" % box_dims[2]) f.write("num_modes = %d\n\n" % num_modes) if exhaustiveness is not None: f.write("exhaustiveness = %d\n" % exhaustiveness) def write_gnina_conf(protein_filename: str, ligand_filename: str, conf_filename: str, num_modes: int = 9, exhaustiveness: int = None, **kwargs) -> None: """Writes GNINA configuration file to disk. GNINA accepts a configuration file which provides options under which GNINA is invoked. This utility function writes a configuration file which directs GNINA to perform docking under the provided options. Parameters ---------- protein_filename: str Filename for protein ligand_filename: str Filename for the ligand conf_filename: str Filename to write Autodock Vina configuration to. num_modes: int, optional (default 9) The number of binding modes GNINA should find exhaustiveness: int, optional The exhaustiveness of the search to be performed by GNINA kwargs: Args supported by GNINA documented here https://github.com/gnina/gnina#usage """ with open(conf_filename, "w") as f: f.write("receptor = %s\n" % protein_filename) f.write("ligand = %s\n\n" % ligand_filename) f.write("autobox_ligand = %s\n\n" % protein_filename) if exhaustiveness is not None: f.write("exhaustiveness = %d\n" % exhaustiveness) f.write("num_modes = %d\n\n" % num_modes) for k, v in kwargs.items(): f.write("%s = %s\n" % (str(k), str(v))) def read_gnina_log(log_file: str) -> np.array: """Read GNINA logfile and get docking scores. GNINA writes computed binding affinities to a logfile. Parameters ---------- log_file: str Filename of logfile generated by GNINA. Returns ------- scores: np.array, dimension (num_modes, 3) Array of binding affinity (kcal/mol), CNN pose score, and CNN affinity for each binding mode. """ scores = [] lines = open(log_file).readlines() mode_start = np.inf for idx, line in enumerate(lines): if line[:6] == '-----+': mode_start = idx if idx > mode_start: mode = line.split() score = [float(x) for x in mode[1:]] scores.append(score) scores = np.array(scores) return scores def load_docked_ligands( pdbqt_output: str) -> Tuple[List[RDKitMol], List[float]]: """This function loads ligands docked by autodock vina. Autodock vina writes outputs to disk in a PDBQT file format. This PDBQT file can contain multiple docked "poses". Recall that a pose is an energetically favorable 3D conformation of a molecule. This utility function reads and loads the structures for multiple poses from vina's output file. Parameters ---------- pdbqt_output: str Should be the filename of a file generated by autodock vina's docking software. Returns ------- Tuple[List[rdkit.Chem.rdchem.Mol], List[float]] Tuple of `molecules, scores`. `molecules` is a list of rdkit molecules with 3D information. `scores` is the associated vina score. Notes ----- This function requires RDKit to be installed. """ try: from rdkit import Chem except ModuleNotFoundError: raise ImportError("This function requires RDKit to be installed.") lines = open(pdbqt_output).readlines() molecule_pdbqts = [] scores = [] current_pdbqt: Optional[List[str]] = None for line in lines: if line[:5] == "MODEL": current_pdbqt = [] elif line[:19] == "REMARK VINA RESULT:": words = line.split() # the line has format # REMARK VINA RESULT: score ... # There is only 1 such line per model so we can append it scores.append(float(words[3])) elif line[:6] == "ENDMDL": molecule_pdbqts.append(current_pdbqt) current_pdbqt = None else: # FIXME: Item "None" of "Optional[List[str]]" has no attribute "append" current_pdbqt.append(line) # type: ignore molecules = [] for pdbqt_data in molecule_pdbqts: pdb_block = pdbqt_to_pdb(pdbqt_data=pdbqt_data) mol = Chem.MolFromPDBBlock(str(pdb_block), sanitize=False, removeHs=False) molecules.append(mol) return molecules, scores def prepare_inputs(protein: str, ligand: str, replace_nonstandard_residues: bool = True, remove_heterogens: bool = True, remove_water: bool = True, add_hydrogens: bool = True, pH: float = 7.0, optimize_ligand: bool = True, pdb_name: Optional[str] = None) -> Tuple[RDKitMol, RDKitMol]: """This prepares protein-ligand complexes for docking. Autodock Vina requires PDB files for proteins and ligands with sensible inputs. This function uses PDBFixer and RDKit to ensure that inputs are reasonable and ready for docking. Default values are given for convenience, but fixing PDB files is complicated and human judgement is required to produce protein structures suitable for docking. Always inspect the results carefully before trying to perform docking. Parameters ---------- protein: str Filename for protein PDB file or a PDBID. ligand: str Either a filename for a ligand PDB file or a SMILES string. replace_nonstandard_residues: bool (default True) Replace nonstandard residues with standard residues. remove_heterogens: bool (default True) Removes residues that are not standard amino acids or nucleotides. remove_water: bool (default True) Remove water molecules. add_hydrogens: bool (default True) Add missing hydrogens at the protonation state given by `pH`. pH: float (default 7.0) Most common form of each residue at given `pH` value is used. optimize_ligand: bool (default True) If True, optimize ligand with RDKit. Required for SMILES inputs. pdb_name: Optional[str] If given, write sanitized protein and ligand to files called "pdb_name.pdb" and "ligand_pdb_name.pdb" Returns ------- Tuple[RDKitMol, RDKitMol] Tuple of `protein_molecule, ligand_molecule` with 3D information. Note ---- This function requires RDKit and OpenMM to be installed. Read more about PDBFixer here: https://github.com/openmm/pdbfixer. Examples -------- >>> p, m = prepare_inputs('3cyx', 'CCC') >>> p.GetNumAtoms() 1415 >>> m.GetNumAtoms() 11 >>> p, m = prepare_inputs('3cyx', 'CCC', remove_heterogens=False) >>> p.GetNumAtoms() 1720 """ try: from rdkit import Chem from rdkit.Chem import AllChem from pdbfixer import PDBFixer from simtk.openmm.app import PDBFile except ModuleNotFoundError: raise ImportError( "This function requires RDKit and OpenMM to be installed.") if protein.endswith('.pdb'): fixer = PDBFixer(protein) else: fixer = PDBFixer(url='https://files.rcsb.org/download/%s.pdb' % (protein)) if ligand.endswith('.pdb'): m = Chem.MolFromPDBFile(ligand) else: m = Chem.MolFromSmiles(ligand, sanitize=True) # Apply common fixes to PDB files if replace_nonstandard_residues: fixer.findMissingResidues() fixer.findNonstandardResidues() fixer.replaceNonstandardResidues() if remove_heterogens and not remove_water: fixer.removeHeterogens(True) if remove_heterogens and remove_water: fixer.removeHeterogens(False) if add_hydrogens: fixer.addMissingHydrogens(pH) PDBFile.writeFile(fixer.topology, fixer.positions, open('tmp.pdb', 'w')) p = Chem.MolFromPDBFile('tmp.pdb', sanitize=True) os.remove('tmp.pdb') # Optimize ligand if optimize_ligand: m = Chem.AddHs(m) # need hydrogens for optimization AllChem.EmbedMolecule(m) AllChem.MMFFOptimizeMolecule(m) if pdb_name: Chem.rdmolfiles.MolToPDBFile(p, '%s.pdb' % (pdb_name)) Chem.rdmolfiles.MolToPDBFile(m, 'ligand_%s.pdb' % (pdb_name)) return (p, m)
deepchem/utils/test/test_vina_utils.py→deepchem/utils/test/test_docking_utils.py +8 −8 Original line number Diff line number Diff line Loading @@ -4,7 +4,7 @@ Test Autodock Vina Utility Functions. import os import numpy as np import unittest from deepchem.utils import vina_utils from deepchem.utils import docking_utils from deepchem.utils import rdkit_utils Loading @@ -17,7 +17,8 @@ class TestVinaUtils(unittest.TestCase): '1jld_ligand_docked.pdbqt') def test_load_docked_ligand(self): docked_ligands, scores = vina_utils.load_docked_ligands(self.docked_ligands) docked_ligands, scores = docking_utils.load_docked_ligands( self.docked_ligands) assert len(docked_ligands) == 9 assert len(scores) == 9 Loading @@ -27,7 +28,7 @@ class TestVinaUtils(unittest.TestCase): assert np.count_nonzero(xyz) > 0 def test_write_gnina_conf(self): vina_utils.write_gnina_conf( docking_utils.write_gnina_conf( 'protein.pdb', 'ligand.sdf', 'conf.txt', Loading @@ -37,7 +38,7 @@ class TestVinaUtils(unittest.TestCase): def test_read_gnina_log(self): log_file = os.path.join(self.current_dir, 'data', 'gnina_log.txt') scores = vina_utils.read_gnina_log(log_file) scores = docking_utils.read_gnina_log(log_file) assert np.array_equal( scores, np.array([[-4.37, 0.6392, 4.336], [-3.56, 0.6202, 4.162]])) Loading @@ -45,13 +46,13 @@ class TestVinaUtils(unittest.TestCase): pdbid = '3cyx' ligand_smiles = 'CC(C)(C)NC(O)C1CC2CCCCC2C[NH+]1CC(O)C(CC1CCCCC1)NC(O)C(CC(N)O)NC(O)C1CCC2CCCCC2N1' protein, ligand = vina_utils.prepare_inputs( protein, ligand = docking_utils.prepare_inputs( pdbid, ligand_smiles, pdb_name=pdbid) assert np.isclose(protein.GetNumAtoms(), 1415, atol=3) assert np.isclose(ligand.GetNumAtoms(), 124, atol=3) protein, ligand = vina_utils.prepare_inputs(pdbid + '.pdb', protein, ligand = docking_utils.prepare_inputs(pdbid + '.pdb', 'ligand_' + pdbid + '.pdb') assert np.isclose(protein.GetNumAtoms(), 1415, atol=3) Loading @@ -59,4 +60,3 @@ class TestVinaUtils(unittest.TestCase): os.remove(pdbid + '.pdb') os.remove('ligand_' + pdbid + '.pdb') os.remove('tmp.pdb')
