Loading deepchem/feat/graph_features.py +86 −0 Original line number Diff line number Diff line Loading @@ -5,7 +5,11 @@ import numpy as np from rdkit import Chem from deepchem.feat import Featurizer from deepchem.feat.atomic_coordinates import ComplexNeighborListFragmentAtomicCoordinates from deepchem.feat.mol_graphs import ConvMol, WeaveMol from deepchem.models.tensorgraph.models.atomic_conv import AtomicConvModel, AtomicConvScore, AtomicConvolution from deepchem.data import DiskDataset import multiprocessing def one_of_k_encoding(x, allowable_set): Loading Loading @@ -424,3 +428,85 @@ class WeaveFeaturizer(Featurizer): graph_distance=self.graph_distance) return WeaveMol(nodes, pairs) class AtomicConvFeaturizer(ComplexNeighborListFragmentAtomicCoordinates): name = ['atomic_conv'] def __init__(self, labels, neighbor_cutoff, frag1_num_atoms=70, frag2_num_atoms=634, complex_num_atoms=701, max_num_neighbors=12, batch_size=24, atom_types=[ 6, 7., 8., 9., 11., 12., 15., 16., 17., 20., 25., 30., 35., 53., -1. ], radial=[[ 1.5, 2.0, 2.5, 3.0, 3.5, 4.0, 4.5, 5.0, 5.5, 6.0, 6.5, 7.0, 7.5, 8.0, 8.5, 9.0, 9.5, 10.0, 10.5, 11.0, 11.5, 12.0 ], [0.0, 4.0, 8.0], [0.4]], layer_sizes=[32, 32, 16], strip_hydrogens=True, learning_rate=0.001, epochs=10): self.atomic_conv_model = AtomicConvModel( frag1_num_atoms=frag1_num_atoms, frag2_num_atoms=frag2_num_atoms, complex_num_atoms=complex_num_atoms, max_num_neighbors=max_num_neighbors, batch_size=batch_size, atom_types=atom_types, radial=radial, layer_sizes=layer_sizes, learning_rate=learning_rate) super(AtomicConvFeaturizer, self).__init__( frag1_num_atoms=frag1_num_atoms, frag2_num_atoms=frag2_num_atoms, complex_num_atoms=complex_num_atoms, max_num_neighbors=max_num_neighbors, neighbor_cutoff=neighbor_cutoff, strip_hydrogens=strip_hydrogens ) self.epochs = epochs self.labels = labels def featurize_complexes(self, mol_files, protein_files): pool = multiprocessing.Pool() results = [] for i, (mol_file, protein_pdb) in enumerate(zip(mol_files, protein_files)): log_message = "Featurizing %d / %d" % (i, len(mol_files)) results.append( pool.apply_async(self._featurize_complex, (self, mol_file, protein_pdb, log_message))) pool.close() features = [] failures = [] for ind, result in enumerate(results): new_features = result.get() # Handle loading failures which return None if new_features is not None: features.append(new_features) else: failures.append(ind) features = np.asarray(features) labels = np.delete(self.labels, failures) dataset = DiskDataset.from_numpy(features, labels) self.atomic_conv_model.fit(dataset, nb_epoch=self.epochs) layers_to_fetch = list() for layer in self.atomic_conv_model.layers.values(): if isinstance(layer, AtomicConvolution) or isinstance(layer, AtomicConvScore): layers_to_fetch.append(layer) atomic_conv_features = list() # TODO (VIGS25): Setup feed_dict and compute the Convolution and score values return atomic_conv_features, failures deepchem/molnet/load_function/pdbbind_datasets.py +17 −0 Original line number Diff line number Diff line Loading @@ -18,6 +18,7 @@ import logging import tarfile from deepchem.feat import rdkit_grid_featurizer as rgf from deepchem.feat.atomic_coordinates import ComplexNeighborListFragmentAtomicCoordinates from deepchem.feat.graph_features import AtomicConvFeaturizer logger = logging.getLogger(__name__) Loading Loading @@ -233,6 +234,22 @@ def load_pdbbind(featurizer="grid", split="random", subset="core", reload=True): frag1_num_atoms, frag2_num_atoms, complex_num_atoms, max_num_neighbors, neighbor_cutoff) elif featurizer == "atomic_conv": frag1_num_atoms = 70 # for ligand atoms frag2_num_atoms = 24000 # for protein atoms complex_num_atoms = 24070 # in total max_num_neighbors = 4 # Cutoff in angstroms neighbor_cutoff = 4 featurizer = AtomicConvFeaturizer( labels=labels, frag1_num_atoms=frag1_num_atoms, frag2_num_atoms=frag2_num_atoms, complex_num_atoms=complex_num_atoms, neighbor_cutoff=neighbor_cutoff, max_num_neighbors=max_num_neighbors, ) else: raise ValueError("Featurizer not supported") print("Featurizing Complexes") Loading Loading
deepchem/feat/graph_features.py +86 −0 Original line number Diff line number Diff line Loading @@ -5,7 +5,11 @@ import numpy as np from rdkit import Chem from deepchem.feat import Featurizer from deepchem.feat.atomic_coordinates import ComplexNeighborListFragmentAtomicCoordinates from deepchem.feat.mol_graphs import ConvMol, WeaveMol from deepchem.models.tensorgraph.models.atomic_conv import AtomicConvModel, AtomicConvScore, AtomicConvolution from deepchem.data import DiskDataset import multiprocessing def one_of_k_encoding(x, allowable_set): Loading Loading @@ -424,3 +428,85 @@ class WeaveFeaturizer(Featurizer): graph_distance=self.graph_distance) return WeaveMol(nodes, pairs) class AtomicConvFeaturizer(ComplexNeighborListFragmentAtomicCoordinates): name = ['atomic_conv'] def __init__(self, labels, neighbor_cutoff, frag1_num_atoms=70, frag2_num_atoms=634, complex_num_atoms=701, max_num_neighbors=12, batch_size=24, atom_types=[ 6, 7., 8., 9., 11., 12., 15., 16., 17., 20., 25., 30., 35., 53., -1. ], radial=[[ 1.5, 2.0, 2.5, 3.0, 3.5, 4.0, 4.5, 5.0, 5.5, 6.0, 6.5, 7.0, 7.5, 8.0, 8.5, 9.0, 9.5, 10.0, 10.5, 11.0, 11.5, 12.0 ], [0.0, 4.0, 8.0], [0.4]], layer_sizes=[32, 32, 16], strip_hydrogens=True, learning_rate=0.001, epochs=10): self.atomic_conv_model = AtomicConvModel( frag1_num_atoms=frag1_num_atoms, frag2_num_atoms=frag2_num_atoms, complex_num_atoms=complex_num_atoms, max_num_neighbors=max_num_neighbors, batch_size=batch_size, atom_types=atom_types, radial=radial, layer_sizes=layer_sizes, learning_rate=learning_rate) super(AtomicConvFeaturizer, self).__init__( frag1_num_atoms=frag1_num_atoms, frag2_num_atoms=frag2_num_atoms, complex_num_atoms=complex_num_atoms, max_num_neighbors=max_num_neighbors, neighbor_cutoff=neighbor_cutoff, strip_hydrogens=strip_hydrogens ) self.epochs = epochs self.labels = labels def featurize_complexes(self, mol_files, protein_files): pool = multiprocessing.Pool() results = [] for i, (mol_file, protein_pdb) in enumerate(zip(mol_files, protein_files)): log_message = "Featurizing %d / %d" % (i, len(mol_files)) results.append( pool.apply_async(self._featurize_complex, (self, mol_file, protein_pdb, log_message))) pool.close() features = [] failures = [] for ind, result in enumerate(results): new_features = result.get() # Handle loading failures which return None if new_features is not None: features.append(new_features) else: failures.append(ind) features = np.asarray(features) labels = np.delete(self.labels, failures) dataset = DiskDataset.from_numpy(features, labels) self.atomic_conv_model.fit(dataset, nb_epoch=self.epochs) layers_to_fetch = list() for layer in self.atomic_conv_model.layers.values(): if isinstance(layer, AtomicConvolution) or isinstance(layer, AtomicConvScore): layers_to_fetch.append(layer) atomic_conv_features = list() # TODO (VIGS25): Setup feed_dict and compute the Convolution and score values return atomic_conv_features, failures
deepchem/molnet/load_function/pdbbind_datasets.py +17 −0 Original line number Diff line number Diff line Loading @@ -18,6 +18,7 @@ import logging import tarfile from deepchem.feat import rdkit_grid_featurizer as rgf from deepchem.feat.atomic_coordinates import ComplexNeighborListFragmentAtomicCoordinates from deepchem.feat.graph_features import AtomicConvFeaturizer logger = logging.getLogger(__name__) Loading Loading @@ -233,6 +234,22 @@ def load_pdbbind(featurizer="grid", split="random", subset="core", reload=True): frag1_num_atoms, frag2_num_atoms, complex_num_atoms, max_num_neighbors, neighbor_cutoff) elif featurizer == "atomic_conv": frag1_num_atoms = 70 # for ligand atoms frag2_num_atoms = 24000 # for protein atoms complex_num_atoms = 24070 # in total max_num_neighbors = 4 # Cutoff in angstroms neighbor_cutoff = 4 featurizer = AtomicConvFeaturizer( labels=labels, frag1_num_atoms=frag1_num_atoms, frag2_num_atoms=frag2_num_atoms, complex_num_atoms=complex_num_atoms, neighbor_cutoff=neighbor_cutoff, max_num_neighbors=max_num_neighbors, ) else: raise ValueError("Featurizer not supported") print("Featurizing Complexes") Loading
