Table 9 Feature summary for each amino acid residue in proteins, each character of SMILES sequence of ligands, and each atom pair of the binding pocket
From: Distance plus attention for binding affinity prediction
Features | Size | Values | Feature description | |
|---|---|---|---|---|
Protein | one-hot encoding | 21 | 1 or 0 | 1 corresponds to the amino acid index, otherwise 0 |
HHM | 30 | real values | various parameters from MSA | |
physicochemical properties | 7 | real values | steric parameter, hydrophobicity, volume, polarisability, isoelectric point, helix probability, sheet probability | |
Ligand | SMILES encoding | 1 | integer values | 64 unique characters, corresponding to a specific numeric digit ranging from 1 to 64 |
Distance bins | 1 | integer values | distances between protein and ligand atoms into 41 bins, with each distance corresponding to a numeric digit from 1 to 41 |