Skip to main content

Table 2 The performance of various SVM models was developed using AC, DC, PSSM profiles and Hybrid methods on the individual Pg-activators SAK, SK, UK and tPA in five-fold cross validation

From: Support vector machine (SVM) based multiclass prediction with basic statistical analysis of plasminogen activators

Proteins

Methods

ACC(%)

SN(%)

SP(%)

MCC

Parameters

      

γ

C

SAK

AC

96.06

92.30

96.91

0.93

3

300

DC

86.82

87.08

86.76

0.83

3

75

PSSM

93.98

92.28

94.34

0.92

1

300

Hybrid

91.72

96.63

90.62

0.93

1

150

SK

AC

95.77

99.05

92.92

0.95

3

275

DC

86.40

93.12

80.56

0.82

10

25

PSSM

97.10

100

94.44

0.97

1

400

Hybrid

90.75

99.05

83.55

0.90

1

250

tPA

AC

95.83

100

95.71

0.97

50

100

DC

92.70

70.58

93.35

0.78

15

450

PSSM

97.73

100

97.67

0.98

4

200

Hybrid

92.69

75.00

93.20

0.81

10

450

UK

AC

90.68

100

86.77

0.93

3

300

DC

87.03

95.03

83.66

0.87

15

500

PSSM

92.06

93.19

91.56

0.90

5

9

Hybrid

85.03

99.40

79.00

0.88

1

450

  1. AC- Amino acid composition, DC dipeptide composition, PSSM position specific scoring matrix, ACC accuracy, SN- Sensitivity, SP- specificity, MCC- Mathews correlation coefficient,C: tradeoff value, γ- gamma factor (a parameter in RBF kernel).
  2. SAK - Staphylokinase, SK - Streptokinase, tPA - tissue plasminogen activators, UK - Urokinase.