Skip to main content

Table 1 10-Fold Cross-Validation performance on six datasets for KNN and SVM-L, λ1=0.5

From: Optimal selection of molecular descriptors for antimicrobial peptides classification: an evolutionary feature weighting approach

Dataset

MLA

Sens(%)

Spec(%)

Prec(%)

Bal Acc(%)

Acc(%)

MCC

AUC

DAMPD_AMP

KNN

71.97

9 7 . 2 2 ∗

8 3 . 7 5 ∗

84.60

9 3 . 0 1

0 . 7 3 5

0.846

 

SVM-L

8 8 . 0 7 ∗a

92.30

69.56

9 0 . 1 9 ∗

91.62

0.734

0.902∗

APD3_AMP

KNN

80.85

95.27∗

7 7 . 2 3 ∗

88.06

9 2 . 8 5

0.747

0.881

 

SVM-L

9 1 . 6 5 ∗

92.53

70.75

9 2 . 0 9 ∗

92.36

0 . 7 6 2

0 . 9 2 1 ∗

DAMPD_ANTIBACTERIAL

KNN

9 1 . 0 4

96.45

8 4 . 3 7

9 3 . 7 5

9 5 . 5 1

0 . 8 4 9

0 . 9 3 7

 

SVM-L

88.49

9 6 . 5 4

84.18

92.51

95.06

0.832

0.925

APD3_ANTIBACTERIAL

KNN

79.32

9 5 . 3 0 ∗

7 7 . 1 8 ∗

87.31

9 2 . 6 1

0.738

0.873

 

SVM-L

9 1 . 3 4 ∗

92.22

70.33

9 1 . 7 8 ∗

92.07

0 . 7 5 6

0 . 9 1 8 ∗

DAMPD_BACTEROCIN

KNN

100

95.53

85.83

97.76

96.36

0.902

0.978

 

SVM-L

100

9 8 . 8 9

9 6 . 6 7

9 9 . 4 4

9 9 . 0 9

0 . 9 7 7

0 . 9 9 4

APD3_BACTEROCIN

KNN

83.50

9 5 . 0 4

77.05

89.27

93.12

0.758

0.893

 

SVM-L

8 5 . 3 8

94.83

7 7 . 2 8

9 0 . 1 0

93.12

0 . 7 6 8

0 . 9 0 1

  1. Each value is the average performance from 10-fold cross-validation by the classifier built by the machine learning algorithm (second column) on the dataset (first column). Wilcoxon signed rank test was performed on the measure resulting from the 10-fold cross-validation of KNN and SVM-L. The models with significant improvement at p-value ≤0.05 are marked with the symbol *
  2. aBold font indicates the best value per measure for every dataset