Skip to main content

Table 3 Comparison among the performances of P2Rank, DeepSurf, PUResNetV1.0, DeepSite, and PUResNetV2.0 on benchmark datasets

From: PUResNetV2.0: a deep learning model leveraging sparse representation for improved ligand binding site prediction

Benchmark dataset

Methods

Success rate (DCA ≤ 4Å) (%)

Success rate (DCC ≤ 4Å) (%)

Coach100

P2Rank

44.0

29.0

DeepSurf

51.0

26.0

PUResNetV1.0

51.0

38.0

DeepSite

27.0

15.0

PUResNetV2.0

59.0

38.0

Holo801

P2Rank

74.0

48.2

DeepSurf

83.8

46.5

PUResNetV1.0

2

0.9

DeepSite

71.4

33.8

PUResNetV2.0

85.4

53.7

Apoholo45

DeepSurf

46.7

16.7

PUResNetV2.0

71.1

40.0

  1. Top values for each benchmark dataset are represented in bold