Table 2 Performance of classifiers before feature selection.

From: XGB-DrugPred: computational prediction of druggable proteins using eXtreme gradient boosting and optimized features set

Classifier

Feature descriptor

Acc (%)

Sn (%)

Sp (%)

F-measure (%)

MCC

ERT

RAAA

81.82

88.10

75.59

82.84

0.64

GDPC

84.65

83.04

85.92

82.67

0.68

S-PseAAC

89.33

88.89

89.76

89.24

0.78

All features

88.14

87.83

88.41

88.69

0.80

RF

RAAA

82.61

86.51

78.74

83.21

0.65

GDPC

83.86

83.93

83.80

82.10

0.67

S-PseAAC

89.72

87.30

92.13

89.43

0.79

All features

90.12

85.22

94.20

88.69

0.80

XGB

RAAA

83.79

84.92

82.95

83.92

0.67

GDPC

86.22

80.36

90.85

83.72

0.72

S-PseAAC

90.51

91.27

89.76

90.55

0.81

All features

92.09

91.30

92.75

91.30

0.84