Skip to main content

Table 2 Prediction performance of different classifiers for vitamin A-interacting residues (VAIRs)

From: Prediction of vitamin interacting residues in a vitamin binding protein using evolutionary information

Feature

    Classifier

SN

SP

ACC

MCC

 Binary

SVM (Threshold = −0.8)

61.92 ± 2.63

65.09 ± 0.43

64.80 ± 0.35

0.16 ± 0.02

SVM (Threshold = −0.1)

7.43 ± 1.18

99.66 ± 0.10

91.28 ± 0.08

0.21 ± 0.02

BayesNet

14.50 ± 2.11

94.30 ± 0.20

87.04 ± 0.22

0.10 ± 0.02

ComplementNaiveBayes

62.09 ± 0.50

65.97 ± 0.22

65.61 ± 0.20

0.17 ± 0.00

NaiveBayes

32.53 ± 0.99

86.43 ± 0.22

81.53 ± 0.27

0.15 ± 0.01

NaiveBayesMultinomial

60.23 ± 0.82

67.94 ± 0.16

67.24 ± 0.15

0.17 ± 0.01

IBk

31.41 ± 2.27

89.80 ± 0.20

84.49 ± 0.19

0.19 ± 0.02

 

RandomForest

36.07 ± 2.03

78.38 ± 0.16

74.54 ± 0.30

0.10 ± 0.01

 PSSM

SVM (Threshold = −0.8)

72.70 ± 2.87

76.89 ± 0.25

76.51 ± 0.37

0.32 ± 0.02

SVM (Threshold =0.0)

42.75 ± 1.08

97.51 ± 0.10

92.54 ± 0.13

0.48 ± 0.01

BayesNet

57.25 ± 1.21

69.54 ± 0.52

68.42 ± 0.48

0.16 ± 0.01

ComplementNaiveBayes

59.30 ± 1.23

66.96 ± 0.33

66.26 ± 0.26

0.16 ± 0.01

NaiveBayes

63.03 ± 1.65

69.09 ± 0.46

68.54 ± 0.56

0.19 ± 0.01

NaiveBayesMultinomial

55.77 ± 1.32

70.95 ± 0.21

69.57 ± 0.26

0.17 ± 0.01

IBk

44.05 ± 0.49

94.65 ± 0.34

90.05 ± 0.27

0.39 ± 0.01

 

RandomForest

24.17 ± 0.80

99.31 ± 0.08

92.49 ± 0.06

0.41 ± 0.01

  1. *Bold value indicates highest performance with balanced sensitivity and specificity.
  2. **Italic value indicates performance with highest MCC.
  3. The values of standard errors are also given with performances.