Skip to main content

Table 9 Comparison of the proposed system with existing nine prediction systems in terms of Serine site prediction in the Independent dataset.

From: Machine learning approach to predict protein phosphorylation sites by incorporating evolutionary information

Systems

Category

Performance parameters of the systems

  

Ac(%)

Sn(%)

Sp(%)

Mcc

FPR(%)

KinasePhos [11]

 

93.11

9.48

94.13

0.02

5.87

NetPhosK [10]

Kinase Specific

85.24

14.22

86.11

0.00

13.89

PPSP [6]

 

80.24

17.54

81.00

-0.00

19.00

GPS [15]

 

80.31

17.54

81.08

-0.00

18.92

AutoMotif Server AMS [14]

 

36.25

64.45

35.90

0.00

64.10

DISPHOS [8]

 

89.40

14.69

90.31

0.02

9.69

NetPhos [13]

 

81.13

16.59

81.91

-0.00

18.09

PHOSIDA [5]

Kinase Independent

94.56

8.53

95.61

0.02

4.39

Scansite [12]

 

98.53

0.95

99.72

0.01

0.28

PPRED (Proposed system)

 

64.96

72.04

64.88

0.08

35.12