Skip to main content

Table 1 The dimensions of features used for atoms and bonds in a molecule graph. Source: Adapted from [29]

From: SSF-DDI: a deep learning method utilizing drug sequence and substructure features for drug–drug interaction prediction

Name

Dimensions

Description

Atom type

Total number of heavy atoms in the dataset

Atom type (e.g., C, O, N)

Degree

11

Count of covalent bonds

Implicit valence

7

Number of implicit H atoms attached to the atom

Hybridization

5

Hybridization rearranges electron orbitals in an atom for efficient covalent bonding (e.g.,sp, sp2, sp3)

Aromatic

1

Whether the atom is situated in an aromatic framework

Formal charge

1

Formal charge of the atom

Radical electrons

1

Number of lone electrons for the atom

Bond type

4

[single, double, triple, aromatic]

Conjugated

1

Whether the bond is involved in a conjugated arrangement

Ring

1

Whether the bond is within a closed loop structure