TABLE 3.

Performance comparison between different methods on the testing set.

Method	Accuracy	Precision	Recall	F-score
Random forest (RF) with amino acid composition (AAC)	0.703 ± 0.012	0.685 ± 0.026	0.703 ± 0.019	0.694 ± 0.022
RF with amino acid pairwise composition (AAPC)	0.711 ± 0.008	0.706 ± 0.017	0.679 ± 0.021	0.692 ± 0.031
RF with k-spaced AAP (k = 5)	0.736 ± 0.006	0.721 ± 0.009	0.714 ± 0.015	0.717 ± 0.019
RF with position-specific scoring matrices (PSMM)	0.722 ± 0.014	0.718 ± 0.008	0.706 ± 0.025	0.713 ± 0.018
Proposed method	0.756 ± 0.006	0.733 ± 0.015	0.767 ± 0.017	0.749 ± 0.009