def test_featurize_binding_pocket_ecfp(self):
prot_xyz, prot_rdk = rgf.load_molecule(self.protein_file)
lig_xyz, lig_rdk = rgf.load_molecule(self.ligand_file)
distance = rgf.compute_pairwise_distances(protein_xyz=prot_xyz,
ligand_xyz=lig_xyz)
# check if results are the same if we provide precomputed distances
prot_dict, lig_dict = rgf.featurize_binding_pocket_ecfp(
prot_xyz,
prot_rdk,
lig_xyz,
lig_rdk,
)
prot_dict_dist, lig_dict_dist = rgf.featurize_binding_pocket_ecfp(
prot_xyz, prot_rdk, lig_xyz, lig_rdk, pairwise_distances=distance)
# ...but first check if we actually got two dicts
self.assertIsInstance(prot_dict, dict)
self.assertIsInstance(lig_dict, dict)
self.assertEqual(prot_dict, prot_dict_dist)
self.assertEqual(lig_dict, lig_dict_dist)
# check if we get less features with smaller distance cutoff
prot_dict_d2, lig_dict_d2 = rgf.featurize_binding_pocket_ecfp(
prot_xyz,
prot_rdk,
lig_xyz,
lig_rdk,
cutoff=2.0,
)
prot_dict_d6, lig_dict_d6 = rgf.featurize_binding_pocket_ecfp(
prot_xyz,
prot_rdk,
lig_xyz,
lig_rdk,
cutoff=6.0,
)
self.assertLess(len(prot_dict_d2), len(prot_dict))
# ligands are typically small so all atoms might be present
self.assertLessEqual(len(lig_dict_d2), len(lig_dict))
self.assertGreater(len(prot_dict_d6), len(prot_dict))
self.assertGreaterEqual(len(lig_dict_d6), len(lig_dict))
# check if using different ecfp_degree changes anything
prot_dict_e3, lig_dict_e3 = rgf.featurize_binding_pocket_ecfp(
prot_xyz,
prot_rdk,
lig_xyz,
lig_rdk,
ecfp_degree=3,
)
self.assertNotEqual(prot_dict_e3, prot_dict)
self.assertNotEqual(lig_dict_e3, lig_dict)
def test_featurize_binding_pocket_ecfp(self):
prot_xyz, prot_rdk = rgf.load_molecule(self.protein_file)
lig_xyz, lig_rdk = rgf.load_molecule(self.ligand_file)
distance = rgf.compute_pairwise_distances(
protein_xyz=prot_xyz, ligand_xyz=lig_xyz)
# check if results are the same if we provide precomputed distances
prot_dict, lig_dict = rgf.featurize_binding_pocket_ecfp(
prot_xyz,
prot_rdk,
lig_xyz,
lig_rdk,
)
prot_dict_dist, lig_dict_dist = rgf.featurize_binding_pocket_ecfp(
prot_xyz, prot_rdk, lig_xyz, lig_rdk, pairwise_distances=distance)
# ...but first check if we actually got two dicts
self.assertIsInstance(prot_dict, dict)
self.assertIsInstance(lig_dict, dict)
self.assertEqual(prot_dict, prot_dict_dist)
self.assertEqual(lig_dict, lig_dict_dist)
# check if we get less features with smaller distance cutoff
prot_dict_d2, lig_dict_d2 = rgf.featurize_binding_pocket_ecfp(
prot_xyz,
prot_rdk,
lig_xyz,
lig_rdk,
cutoff=2.0,
)
prot_dict_d6, lig_dict_d6 = rgf.featurize_binding_pocket_ecfp(
prot_xyz,
prot_rdk,
lig_xyz,
lig_rdk,
cutoff=6.0,
)
self.assertLess(len(prot_dict_d2), len(prot_dict))
# ligands are typically small so all atoms might be present
self.assertLessEqual(len(lig_dict_d2), len(lig_dict))
self.assertGreater(len(prot_dict_d6), len(prot_dict))
self.assertGreaterEqual(len(lig_dict_d6), len(lig_dict))
# check if using different ecfp_degree changes anything
prot_dict_e3, lig_dict_e3 = rgf.featurize_binding_pocket_ecfp(
prot_xyz,
prot_rdk,
lig_xyz,
lig_rdk,
ecfp_degree=3,
)
self.assertNotEqual(prot_dict_e3, prot_dict)
self.assertNotEqual(lig_dict_e3, lig_dict)
def test_voxelize(self):
prot_xyz, prot_rdk = rgf.load_molecule(self.protein_file)
lig_xyz, lig_rdk = rgf.load_molecule(self.ligand_file)
centroid = rgf.compute_centroid(lig_xyz)
prot_xyz = rgf.subtract_centroid(prot_xyz, centroid)
lig_xyz = rgf.subtract_centroid(lig_xyz, centroid)
prot_ecfp_dict, lig_ecfp_dict = rgf.featurize_binding_pocket_ecfp(
prot_xyz, prot_rdk, lig_xyz, lig_rdk)
box_w = 20
f_power = 5
rgf_featurizer = rgf.RdkitGridFeaturizer(
box_width=box_w,
ecfp_power=f_power,
feature_types=['all_combined'],
flatten=True,
sanitize=True)
prot_tensor = rgf_featurizer._voxelize(
rgf.convert_atom_to_voxel,
rgf.hash_ecfp,
prot_xyz,
feature_dict=prot_ecfp_dict,
channel_power=f_power)
self.assertEqual(prot_tensor.shape, tuple([box_w] * 3 + [2**f_power]))
all_features = prot_tensor.sum()
# protein is too big for the box, some features should be missing
self.assertGreater(all_features, 0)
self.assertLess(all_features, prot_rdk.GetNumAtoms())
lig_tensor = rgf_featurizer._voxelize(
rgf.convert_atom_to_voxel,
rgf.hash_ecfp,
lig_xyz,
feature_dict=lig_ecfp_dict,
channel_power=f_power)
self.assertEqual(lig_tensor.shape, tuple([box_w] * 3 + [2**f_power]))
all_features = lig_tensor.sum()
# whole ligand should fit in the box
self.assertEqual(all_features, lig_rdk.GetNumAtoms())
def test_voxelize(self):
prot_xyz, prot_rdk = rgf.load_molecule(self.protein_file)
lig_xyz, lig_rdk = rgf.load_molecule(self.ligand_file)
centroid = rgf.compute_centroid(lig_xyz)
prot_xyz = rgf.subtract_centroid(prot_xyz, centroid)
lig_xyz = rgf.subtract_centroid(lig_xyz, centroid)
prot_ecfp_dict, lig_ecfp_dict = rgf.featurize_binding_pocket_ecfp(
prot_xyz, prot_rdk, lig_xyz, lig_rdk)
box_w = 20
f_power = 5
rgf_featurizer = rgf.RdkitGridFeaturizer(
box_width=box_w,
ecfp_power=f_power,
feature_types=['all_combined'],
flatten=True,
sanitize=True)
prot_tensor = rgf_featurizer._voxelize(
rgf.convert_atom_to_voxel,
rgf.hash_ecfp,
prot_xyz,
feature_dict=prot_ecfp_dict,
channel_power=f_power)
self.assertEqual(prot_tensor.shape, tuple([box_w] * 3 + [2**f_power]))
all_features = prot_tensor.sum()
# protein is too big for the box, some features should be missing
self.assertGreater(all_features, 0)
self.assertLess(all_features, prot_rdk.GetNumAtoms())
lig_tensor = rgf_featurizer._voxelize(
rgf.convert_atom_to_voxel,
rgf.hash_ecfp,
lig_xyz,
feature_dict=lig_ecfp_dict,
channel_power=f_power)
self.assertEqual(lig_tensor.shape, tuple([box_w] * 3 + [2**f_power]))
all_features = lig_tensor.sum()
# whole ligand should fit in the box
self.assertEqual(all_features, lig_rdk.GetNumAtoms())