Skip to main content

Table 3 Comparison of ECMIS with other prediction methods

From: ECMIS: computational approach for the identification of hotspots at protein-protein interfaces

Method

PR (%)

SE (%)

SP (%)

ACC (%)

MCC

ECMIS

68%

66%

87%

80%

0.524

RF

70.8

44.7

92.0

77.6

0.429

MINERVA

65.4

44.7

89.7

76.2

0.390

KFC2

58.1

47.4

85.1

73.6

0.345

HotPoint

49.0

63.2

71.3

68.8

0.324

Robetta

52.0

34.2

86.2

70.4

0.235

KFC

48.0

31.6

85.1

68.8

0.191

FOLDEF

47.6

26.3

87.4

68.8

0.168