Skip to main content

Table 1 The performance of SVM models using amino acid compositions of whole peptides and five residues of N-terminus. Last column shows performance of binary-based SVM model using first five residues of peptides.

From: Analysis and prediction of antibacterial peptides

 

Whole Peptide Composition

5 N-terminal residue Composition

5 N-terminal binary Pattern

Threshold

Sen.

Spec.

Acc.

Sen.

Spec.

Acc.

Sen.

Spec.

Acc.

-1

98.15

60.45

80.13

97.31

25.45

61.38

95.81

24.25

60.03

-0.9

97.94

63.60

81.53

95.21

30.54

62.87

94.31

30.54

62.43

-0.8

97.53

67.19

83.03

93.71

35.33

64.52

93.71

37.13

65.42

-0.7

97.12

70.56

84.43

91.62

39.82

65.72

91.62

43.11

67.37

-0.6

96.91

72.81

85.39

90.42

46.11

68.26

90.12

48.50

69.31

-0.5

96.30

76.18

86.68

89.22

50.60

69.91

86.83

54.19

70.51

-0.4

95.88

77.53

87.11

87.72

55.09

71.41

85.03

59.88

72.46

-0.3

94.65

80.90

88.08

84.73

58.98

71.86

82.34

64.97

73.65

-0.2

93.42

84.27

89.04

81.44

63.47

72.46

80.24

68.86

74.55

0

89.92

88.09

89.04

73.65

72.75

73.20

74.25

74.85

74.55

0.1

88.48

89.44

88.94

70.36

76.65

73.50

70.66

77.54

74.10

0.2

86.63

90.56

88.51

66.77

79.64

73.20

67.96

83.83

75.90

0.3

84.98

91.01

87.86

63.47

83.83

73.65

64.97

88.32

76.65

0.4

83.74

93.03

88.18

57.49

86.83

72.16

60.78

88.92

74.85

0.5

80.04

93.26

86.36

53.59

89.52

71.56

57.78

90.12

73.95

0.6

77.37

94.83

85.71

48.50

91.62

70.06

54.49

91.02

72.75

0.7

75.51

95.28

84.96

44.61

94.01

69.31

51.50

93.41

72.46

0.8

72.02

95.28

83.14

39.22

94.91

67.07

44.91

95.51

70.21

0.9

67.90

96.40

81.53

34.43

96.11

65.27

39.52

96.11

67.81

1

64.61

97.53

80.34

29.04

97.90

63.47

32.34

97.01

64.67