Table 2 Evaluation performance metrics of RF classifier model on independent dataset using various features derived from Pfeature

From: Prediction of hemolytic peptides and their hemolytic concentration

Features

Sp (%)

Sn (%)

Acc (%)

MCC

AUC

AAC

83.9

66.3

75.9

0.512

0.831

DPC

87.7

65.1

77.5

0.547

0.837

PCP

84.8

67.4

76.9

0.534

0.854

CeTD

86.7

72.0

80.1

0.597

0.856

AAC + DPC

74.9

85.5

80.6

0.609

0.875

AAC + DPC+CeTD

76.2

78.8

77.5

0.598

0.841

AAC + DPC + PCP

73.7

85.0

79.8

0.623

0.881

ALLCOMP

74.9

83.9

80.1

0.610

0.878

ALLCOMP-ex SOC

76.5

86.9

82.1

0.640

0.888

  1. Sn sensitivity, Sp specificity, Acc accuracy, MCC Matthews correlation coefficient, AUC area under receiver operating characteristic, Bold values indicate the best-performing model.