dgllife.utils.AttentiveFPAtomFeaturizer

class dgllife.utils.AttentiveFPAtomFeaturizer(atom_data_field='h')[source]

The atom featurizer used in AttentiveFP

AttentiveFP is introduced in Pushing the Boundaries of Molecular Representation for Drug Discovery with the Graph Attention Mechanism.

The atom features include:

  • One hot encoding of the atom type. The supported atom types include B, C, N, O, F, Si, P, S, Cl, As, Se, Br, Te, I, At, and other.

  • One hot encoding of the atom degree. The supported possibilities include 0 - 5.

  • Formal charge of the atom.

  • Number of radical electrons of the atom.

  • One hot encoding of the atom hybridization. The supported possibilities include SP, SP2, SP3, SP3D, SP3D2, and other.

  • Whether the atom is aromatic.

  • One hot encoding of the number of total Hs on the atom. The supported possibilities include 0 - 4.

  • Whether the atom is chiral center

  • One hot encoding of the atom chirality type. The supported possibilities include R, and S.

We assume the resulting DGLGraph will not contain any virtual nodes.

Parameters

atom_data_field (str) – Name for storing atom features in DGLGraphs, default to ‘h’.

Examples

>>> from rdkit import Chem
>>> from dgllife.utils import AttentiveFPAtomFeaturizer

>>> mol = Chem.MolFromSmiles('CCO')
>>> atom_featurizer = AttentiveFPAtomFeaturizer(atom_data_field='feat')
>>> atom_featurizer(mol)
{'feat': tensor([[0., 1., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 1.,
                  0., 0., 0., 0., 0., 0., 0., 0., 1., 0., 0., 0., 0., 0., 0., 0., 1., 0.,
                  0., 0., 0.],
                 [0., 1., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0.,
                  1., 0., 0., 0., 0., 0., 0., 0., 1., 0., 0., 0., 0., 0., 0., 1., 0., 0.,
                  0., 0., 0.],
                 [0., 0., 0., 1., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 0., 1.,
                  0., 0., 0., 0., 0., 0., 0., 0., 1., 0., 0., 0., 0., 0., 1., 0., 0., 0.,
                  0., 0., 0.]])}

>>> # Get feature size for nodes
>>> print(atom_featurizer.feat_size('feat'))
39

See also

BaseAtomFeaturizer, CanonicalAtomFeaturizer, WeaveAtomFeaturizer, PretrainAtomFeaturizer, PAGTNAtomFeaturizer

__init__(atom_data_field='h')[source]

Initialize self. See help(type(self)) for accurate signature.

Methods

__init__([atom_data_field])

Initialize self.

feat_size([feat_name])

Get the feature size for feat_name.