Skip to main content

Table 3 Combination of different feature groups and performance comparison based on 5-fold cross-validation tests on ATP168 and ATP227

From: Predicting protein-ATP binding sites from primary sequence through fusing bi-profile sampling of multi-view features

 

Composition of different features

Sen (%)

Spe (%)

Acc (%)

AUC

ATP168

PSSM

75.67

75.71

75.82

0.8410

LogisticPSSMa

76.28

76.62

76.60

0.8493

LogisticPSSM + Bipro-aa

76.90

77.19

77.18

0.8553

LogisticPSSM + Bipro-dis

76.64

77.09

77.07

0.8537

LogisticPSSM + Bipro-sa

77.00

77.48

77.46

0.8562

LogisticPSSM + Bipro-ss

77.35

77.58

77.56

0.8579

LogisticPSSM + Bipro-allb

77.00

77.32

77.30

0.8569

ATP227

PSSM

78.31

78.16

78.16

0.8609

LogisticPSSMa

79.17

79.59

79.57

0.8727

LogisticPSSM + Bipro-aa

79.77

79.91

79.91

0.8770

LogisticPSSM + Bipro-dis

79.71

79.89

79.88

0.8763

LogisticPSSM + Bipro-sa

79.95

80.12

80.11

0.8797

LogisticPSSM + Bipro-ss

80.07

80.31

80.30

0.8813

 

LogisticPSSM + Bipro-allb

79.95

80.27

80.26

0.8800

  1. a LogisticPSSM means PSSM profile normalized by logistic function.
  2. b all means the combination of four feature groups: Bipro-aa, Bipro-dis, Bipro-sa, and Bipro-ss.