DIRProt: a computational approach for discriminating insecticide resistant proteins from non-resistant proteins

BMC Bioinformatics

Table 2 Estimates of performance metrics for classification of detoxification and target-based resistant proteins, under different feature sets

Feature	Sn	Sp	Ac	Pre	MCC	AUC-ROC
AAC	0.927 ± 0.020	0.966 ± 0.042	0.946 ± 0.024	0.966 ± 0.041	0.894 ± 0.049	0.960 ± 0.023
DPC	0.967 ± 0.067	0.985 ± 0.031	0.976 ± 0.035	0.986 ± 0.029	0.955 ± 0.065	0.972 ± 0.051
PAAC	0.929 ± 0.016	0.952 ± 0.048	0.941 ± 0.027	0.953 ± 0.046	0.883 ± 0.054	0.956 ± 0.028
CTD	0.895 ± 0.042	0.979 ± 0.035	0.937 ± 0.024	0.979 ± 0.035	0.879 ± 0.047	0.935 ± 0.036
ACF	0.912 ± 0.041	0.927 ± 0.051	0.919 ± 0.037	0.927 ± 0.049	0.840 ± 0.074	0.967 ± 0.021

Sn Sensitivity, Sp Specificity, Ac Accuracy, Pre Precision, MCC Matthew’s correlation coefficient, AUC-ROC area under ROC curves

ISSN: 1471-2105