Skip to main content

Table 4 Overview of the prediction accuracy of the Large Dataset (404 Residues) II Three tables show the accuracy of the predictions to the measured pKexp within the ranges of <2 to <0.5. This is taken as the number of residues predicted within each range. Figure marked in bold indicate significant results (P = 0.05).

From: Benchmarking pKa prediction

TOTAL (404)

        
 

AMBER

 

PARSE

 

MCCE

 

PROPKA

 
  

%

 

%

 

%

 

%

<2

322

80

320

79

357

88

371

92

<1.5

283

70

282

70

328

81

354

88

<1

213

53

221

55

283

70

317

78

<0.5

108

27

139

34

168

42

195

48

SURFACE (337)

        
 

AMBER

 

PARSE

 

MCCE

 

PROPKA

 
  

%

 

%

 

%

 

%

<2

285

85

284

84

311

92

329

98

<1.5

255

76

253

75

290

86

317

94

<1

196

58

200

59

248

74

285

85

<0.5

98

29

132

39

151

45

184

55

BURIED (66)

        
 

AMBER

 

PARSE

 

MCCE

 

PROPKA

 
  

%

 

%

 

%

 

%

<2

37

56

36

55

46

70

42

64

<1.5

28

42

29

44

38

58

37

56

<1

17

26

21

32

35

53

32

48

<0.5

10

15

7

11

17

26

11

17