Skip to main content

Table 3 Comparison with three classifier on four datasets in term of ACC, TPR, SPC, PPV, precision, MCC, and AUC

From: A deep learning-based method for drug-target interaction prediction based on long short-term memory neural network

Datasets

Model

ACC (%)

TPR (%)

SPC (%)

PPV (%)

MCC (%)

AUC

enzymes

MLP

90.01

100

80.06

83.31

81.67

0.9967

SVM

89.88

92.31

87.53

88.12

81.77

0.9686

DeepLSTM

92.92

99.31

86.57

88.04

86.75

0.9951

ion channels

MLP

87.58

100

75.22

80.07

77.61

0.9972

SVM

89.36

85.95

92.74

92.23

80.93

0.9613

DeepLSTM

91.97

93.23

90.87

89.95

85.19

0.9705

GPCRs

MLP

87.20

76.70

97.77

97.19

77.20

0.9853

SVM

85.43

86.28

84.60

84.81

74.99

0.9230

DeepLSTM

91.80

83.71

100

100

84.44

0.9951

nucl. rec

MLP

88.89

88.24

89.47

88.24

80.19

0.8421

SVM

85.00

68.90

100

100

72.43

0.9910

DeepLSTM

91.11

95.24

87.50

86.96

83.76

0.9206