Skip to main content

Table 1 Comparison of docking performances between MOE and Maestro data preparation schemes for diverse protein targets

From: Applying DEKOIS 2.0 in structure-based virtual screening to probe the impact of preparation procedures and score normalization

Target

Glide

GOLD

Original pROC-AUC

Original pROC-AUC

MOE

Maestro

ΔpROC-AUC prep a

MOE

Maestro

ΔpROC-AUC prep

ACE

1.73

2.01

−0.28 b

1.55

1.25

0.30

ACHE

0.71

0.75

−0.04

0.72

0.72

0.00

ADRB2

0.97

0.89

0.09

0.74

0.65

0.08

CATL

0.72

0.81

−0.09

0.78

0.75

0.04

DHFR

1.12

0.63

0.49 c

0.75

0.80

−0.05

ERBB2

2.01

1.80

0.21

1.85

2.12

−0.27

HDAC2

1.74

1.27

0.47

1.02

1.20

−0.18

HIVPR

1.54

1.40

0.13

1.46

1.64

−0.18

HSP90

0.74

0.61

0.14

0.42

0.33

0.10

JAK3

1.32

1.13

0.19

0.78

0.91

−0.14

JNK2

0.66

0.70

−0.04

0.77

0.74

0.03

MDM2

0.84

0.80

0.04

0.51

0.44

0.07

P38

1.79

1.48

0.31

0.49

0.56

−0.07

PI3KG

1.14

1.13

0.01

1.01

0.95

0.06

PNP

1.08

1.41

−0.33

0.98

1.04

−0.07

PPARg

0.90

1.11

−0.20

0.85

0.92

−0.07

Thrombin

2.00

1.93

0.07

1.22

1.15

0.08

TS

1.22

1.48

−0.26

1.27

1.05

0.22

  1. aΔpROC-AUC prep is calculated as pROC-AUC (MOE) – pROC-AUC (Maestro). bItalicized values represent significant deviations (ΔpROC-AUC prep ), while the non-italicized values are deemed non-significant. cBold values represent the maximum ΔpROC-AUC prep observed (for GOLD: ACE; for Glide: DHFR)