Skip to main content

Table 4 Performance of various AGL-EAT-score models on CASF–2016 test set

From: The algebraic extended atom-type graph-based model for precise ligand–receptor binding affinity prediction

Model

Trained with refined set

Trained with general set

RMSE

MAE

SD

R

RMSE

MAE

SD

R

\(\text {AGL-EAT}_{\Phi _E,16.5,3.0}^{\textrm{Adj}}\)

1.29

1.02

0.78

0.835

1.15

0.89

0.72

0.873

\(\text {AGL-EAT}_{\Phi _E,19.5,2.5}^{\textrm{Lap}}\)

1.28

1.02

0.78

0.837

1.15

0.90

0.72

0.871

  1. Error metrics are in pKd unit