Skip to main content

Table 1 Performance comparisons on PDBbind and the modified Davis sets

From: Multi-PLI: interpretable multi‐task deep learning model for unifying protein–ligand interaction datasets

Dataset

Method

RMSE

R

SD

PDBbind (core2016)

SVM

1.77

0.48

1.68

DeepDTA

1.51

0.61

1.50

 

Pafnucy

1.42

0.78

1.37

 

Single-task on PDBbind

1.44

0.75

1.44

 

Single-task on Davis

2.51

0.32

2.06

 

Multi-task

1.49

0.74

1.46

Davis

SVM

0.98

0.36

0.96

 

DeepDTA

0.95

0.43

0.94

 

Single-task on Davis

0.92

0.46

0.92

 

Single-task on PDBbind

1.26

0.03

1.06

 

Multi-task

0.99

0.38

0.97

  1. The result of Pafnucy on PDBbind is derived from [11]
  2. Best values are higlhlighted in bold
Back to article page

Journal of Cheminformatics

ISSN: 1758-2946

Contact us