Skip to main content

Table 3 Comparison of IUPforest-L with other predictors on test set Peng-DB.*

From: Large-scale prediction of long disordered regions in proteins using random forests

Measure(%)

sen

spe

MCC

Spro

Bacc

Sw

AUC

IUPforest-L

39.1

98.3

24.8

38.4

68.7

41.0

83.2

DisEMBL

43.2

89.9

10.1

38.8

66.5

33.1

73.7

DISOPRED2

40.9

96.7

19.3

39.5

68.8

39.6

80.0

FoldIndex

49.1

84.4

85.8

41.4

66.7

35.7

N/A

FoldUnfold

37.9

90.9

31.0

34.5

64.4

19.0

N/A

IUPred

43.2

92.8

12.8

40.1

68.0

36.1

75.2

POODLE-L

39.9

96.8

18.6

38.9

68.3

36.7

79.8

RONN

42.3

96.3

18.3

40.7

69.3

38.6

79.2

Spritz (long)

23.1

96.9

18.6

22.4

60.0

-19

N/A

VL3H

40.2

96.8

11.8

38.9

68.5

65.6

91.3

VSL2

58.2

92.9

14.6

54.1

75.6

61.0

85.4

VSL2B

42.2

95.7

16.9

40.4

69.0

38.0

78.9

  1. *A random sample of 112 sequences from the ordered test set of Peng-DB was used to test Spritz(long) and DISOPRED2 due to the size restriction of the prediction services. N/A means no raw score was available for comparison.