more cleanup (38d261f0) · Commits · 钟慕尧 / deepchem

examples/membrane_permeability/membrane_permeability_datasets.py→deepchem/molnet/load_function/membrane_permeability_datasets.py

+15 −6

Original line number	Diff line number	Diff line
		"""
		MUV dataset loader.
		Membrane Permeability Dataset Loader
		"""
		from __future__ import print_function
		from __future__ import division
		from __future__ import unicode_literals

		import os
		import numpy as np
		import shutil
		@@ -19,7 +15,11 @@ def load_permeability(featurizer='ECFP', split='index'):
		Leung, Siegfried SF, Daniel Sindhikara, and Matthew P. Jacobson. "Simple predictive models of passive membrane permeability incorporating size-dependent membrane-water partition." Journal of chemical information and modeling 56.5 (2016): 924-929.


		Looks like we're using the 3d coordinates here. Just like ESOL, permeability should be based on the compound not conformer, but the conformational ensemble highly affects the permeability (solubility as well). Existing predictors of permeability and solubility both require sampling of the 3d structures.
		Just like the ESOL dataset, permeability should be based on
		the compound not conformer, but the conformational ensemble
		highly affects the permeability (solubility as well).
		Existing predictors of permeability and solubility both often
		require sampling of the 3d structures.
		"""
		print("About to load membrane permeability dataset.")
		current_dir = os.path.dirname(os.path.realpath(__file__))
		@@ -32,6 +32,15 @@ def load_permeability(featurizer='ECFP', split='index'):
		featurizer_func = dc.feat.CircularFingerprint(size=1024)
		elif featurizer == 'GraphConv':
		featurizer_func = dc.feat.ConvMolFeaturizer()
		elif featurizer == 'Weave':
		featurizer = deepchem.feat.WeaveFeaturizer()
		elif featurizer == 'Raw':
		featurizer = deepchem.feat.RawFeaturizer()
		elif featurizer == "smiles2img":
		img_spec = kwargs.get("img_spec", "std")
		img_size = kwargs.get("img_size", 80)
		featurizer = deepchem.feat.SmilesToImage(
		img_size=img_size, img_spec=img_spec)

		permeability_tasks = sorted(['LogP(RRCK)'])

examples/delaney/delaney_graph_conv.py

+0 −4

Original line number	Diff line number	Diff line
		"""
		Script that trains graph-conv models on Tox21 dataset.
		"""
		from __future__ import print_function
		from __future__ import division
		from __future__ import unicode_literals

		import numpy as np

		from deepchem.models import GraphConvModel

examples/delaney/delaney_graphconv_error_bars.py

+0 −4

Original line number	Diff line number	Diff line
		"""
		Script that trains graphconv models on delaney dataset.
		"""
		from __future__ import print_function
		from __future__ import division
		from __future__ import unicode_literals

		import numpy as np
		np.random.seed(123)
		import tensorflow as tf

examples/hopv/hopv_tf_models.py→examples/hopv/hopv_fcnet.py

+0 −0

File moved.

View file

examples/hopv/hopv_graph_conv.py

+0 −4

Original line number	Diff line number	Diff line
		"""
		Script that trains graph-conv models on HOPV dataset.
		"""
		from __future__ import print_function
		from __future__ import division
		from __future__ import unicode_literals

		import numpy as np

		from deepchem.models import GraphConvModel

Admin message