Table 5. The performance of our method and other existing methods on PDB594 dataset (Jackknife test evaluation).

Methods	ACC(%)	MCC	SN(%)	SP(%)
GNB-based-wrapper	80.5	0.610	82.8	78.1
DT-based-wrapper	69.2	0.384	68.4	70.0
LongR-based-wrapper	75.4	0.511	80.5	70.4
KNN-based-wrapper	74.6	0.492	72.1	77.1
SVM-Poly-based-wrapper	77.1	0.550	85.5	68.7
SVM-RBF-based-wrapper	80.1	0.605	84.8	75.4
Our method	86.2	0.724	87.2	85.2