:recycle: refactor base featurizer classes (d7a99002) · Commits · 钟慕尧 / deepchem

deepchem/dock/docking.py

+1 −1

Original line number	Diff line number	Diff line
		@@ -109,7 +109,7 @@ class Docker(object):
		if self.scoring_model is not None:
		for posed_complex in complexes:
		# TODO: How to handle the failure here?
		features, _ = self.featurizer.featurize_complexes([molecular_complex])
		features, _ = self.featurizer.featurize([molecular_complex])
		dataset = NumpyDataset(X=features)
		score = self.scoring_model.predict(dataset)
		yield (posed_complex, score)

deepchem/dock/tests/test_docking.py

+1 −1

Original line number	Diff line number	Diff line
		@@ -105,7 +105,7 @@ class TestDocking(unittest.TestCase):

		class DummyFeaturizer(ComplexFeaturizer):

		def featurize_complexes(self, complexes, args, *kwargs):
		def featurize(self, complexes, args, *kwargs):
		return np.zeros((len(complexes), 5)), None

		class DummyModel(Model):

deepchem/feat/atomic_coordinates.py

+2 −3

Original line number	Diff line number	Diff line
		@@ -3,7 +3,6 @@ Atomic coordinate featurizer.
		"""
		import logging
		import numpy as np
		from deepchem.utils.save import log
		from deepchem.feat import Featurizer
		from deepchem.feat import ComplexFeaturizer
		from deepchem.utils import rdkit_util, pad_array
		@@ -162,7 +161,7 @@ class NeighborListComplexAtomicCoordinates(ComplexFeaturizer):
		self.dtype = object
		self.coordinates_featurizer = AtomicCoordinates()

		def _featurize_complex(self, mol_pdb_file, protein_pdb_file):
		def _featurize(self, mol_pdb_file, protein_pdb_file):
		"""
		Compute neighbor list for complex.

		@@ -218,7 +217,7 @@ class ComplexNeighborListFragmentAtomicCoordinates(ComplexFeaturizer):
		self.neighborlist_featurizer = NeighborListComplexAtomicCoordinates(
		self.max_num_neighbors, self.neighbor_cutoff)

		def _featurize_complex(self, mol_pdb_file, protein_pdb_file):
		def _featurize(self, mol_pdb_file, protein_pdb_file):
		try:
		frag1_coords, frag1_mol = rdkit_util.load_molecule(
		mol_pdb_file, is_protein=False, sanitize=True, add_hydrogens=False)

deepchem/feat/base_classes.py

+12 −88

Original line number	Diff line number	Diff line
		@@ -12,11 +12,6 @@ logger = logging.getLogger(__name__)
		JSON = Dict[str, Any]


		def _featurize_complex(featurizer, mol_pdb_file, protein_pdb_file, log_message):
		logging.info(log_message)
		return featurizer._featurize_complex(mol_pdb_file, protein_pdb_file)


		class Featurizer(object):
		"""Abstract class for calculating a set of features for a datapoint.

		@@ -57,24 +52,23 @@ class Featurizer(object):
		features = np.asarray(features)
		return features

		def __call__(self, datapoints):
		"""Calculate features for datapoints.
		def _featurize(self, datapoint):
		"""Calculate features for a single datapoint.

		Parameters
		----------
		datapoints: object
		Any blob of data you like. Subclasss should instantiate
		this.
		datapoint: object
		a single datapoint in a sequence of objects
		"""
		return self.featurize(datapoints)
		raise NotImplementedError('Featurizer is not defined.')


		class ComplexFeaturizer(object):
		class ComplexFeaturizer(Featurizer):
		""""
		Abstract class for calculating features for mol/protein complexes.
		"""

		def featurize_complexes(self, mol_files, protein_pdbs):
		def featurize(self, mol_files, protein_pdbs):
		"""
		Calculate features for mol/protein complexes.

		@@ -97,7 +91,7 @@ class ComplexFeaturizer(object):
		for i, (mol_file, protein_pdb) in enumerate(zip(mol_files, protein_pdbs)):
		log_message = "Featurizing %d / %d" % (i, len(mol_files))
		results.append(
		pool.apply_async(_featurize_complex,
		pool.apply_async(self._featurize,
		(self, mol_file, protein_pdb, log_message)))
		pool.close()
		features = []
		@@ -112,7 +106,7 @@ class ComplexFeaturizer(object):
		features = np.asarray(features)
		return features, failures

		def _featurize_complex(self, mol_pdb, complex_pdb):
		def _featurize(self, mol_pdb, complex_pdb):
		"""
		Calculate features for single mol/protein complex.

		@@ -187,28 +181,6 @@ class MolecularFeaturizer(Featurizer):
		features = np.asarray(features)
		return features

		def _featurize(self, mol):
		"""
		Calculate features for a single molecule.

		Parameters
		----------
		mol : RDKit Mol
		Molecule.
		"""
		raise NotImplementedError('Featurizer is not defined.')

		def __call__(self, molecules):
		"""
		Calculate features for molecules.

		Parameters
		----------
		molecules: iterable
		An iterable yielding RDKit Mol objects or SMILES strings.
		"""
		return self.featurize(molecules)


		class StructureFeaturizer(Featurizer):
		"""
		@@ -282,30 +254,6 @@ class StructureFeaturizer(Featurizer):
		features = np.asarray(features)
		return features

		def _featurize(self, structure: "pymatgen.Structure"):
		"""Calculate features for a single crystal structure.

		Parameters
		----------
		structure: pymatgen.Structure object
		Structure object with 3D coordinates and periodic lattice.

		"""

		raise NotImplementedError('Featurizer is not defined.')

		def __call__(self, structures: Iterable[dict]):
		"""Calculate features for crystal structures.

		Parameters
		----------
		structures: Iterable[dict]
		An iterable of crystal structure dictionaries.

		"""

		return self.featurize(structures)


		class CompositionFeaturizer(Featurizer):
		"""
		@@ -377,30 +325,6 @@ class CompositionFeaturizer(Featurizer):
		features = np.asarray(features)
		return features

		def _featurize(self, composition: "pymatgen.Composition"):
		"""Calculate features for a single crystal composition.

		Parameters
		----------
		composition: pymatgen.Composition object
		Composition object for 3D inorganic crystal.

		"""

		raise NotImplementedError('Featurizer is not defined.')

		def __call__(self, compositions: Iterable[str]):
		"""Calculate features for crystal compositions.

		Parameters
		----------
		compositions: Iterable[str]
		An iterable of crystal compositions.

		"""

		return self.featurize(compositions)


		class UserDefinedFeaturizer(Featurizer):
		"""Directs usage of user-computed featurizations."""

deepchem/feat/graph_features.py

+2 −9

Original line number	Diff line number	Diff line
		import enum
		import numpy as np
		import deepchem as dc
		from deepchem.feat.base_classes import MolecularFeaturizer
		from deepchem.feat.atomic_coordinates import ComplexNeighborListFragmentAtomicCoordinates
		from deepchem.feat.mol_graphs import ConvMol, WeaveMol
		from deepchem.data import DiskDataset
		import multiprocessing
		import logging


		def _featurize_complex(featurizer, mol_pdb_file, protein_pdb_file, log_message):
		logging.info(log_message)
		return featurizer._featurize_complex(mol_pdb_file, protein_pdb_file)


		def one_of_k_encoding(x, allowable_set):
		"""Encodes elements of a provided set as integers.

		@@ -815,12 +808,12 @@ class AtomicConvFeaturizer(ComplexNeighborListFragmentAtomicCoordinates):
		self.epochs = epochs
		self.labels = labels

		def featurize_complexes(self, mol_files, protein_files):
		def featurize(self, mol_files, protein_files):
		features = []
		failures = []
		for i, (mol_file, protein_pdb) in enumerate(zip(mol_files, protein_files)):
		logging.info("Featurizing %d / %d" % (i, len(mol_files)))
		new_features = self._featurize_complex(mol_file, protein_pdb)
		new_features = self._featurize(mol_file, protein_pdb)
		# Handle loading failures which return None
		if new_features is not None:
		features.append(new_features)

Admin message