Table 4. Performance of models based on direct feature combination for the unbound sequence dataset, evaluated by LOOCV.

Feature	F	SN	SP	ACC	AUC
A	0.292	0.636	0.643	0.654	0.635
A+B	0.288	0.628	0.653	0.658	0.634
A+B+C	0.283	0.644	0.636	0.646	0.633
A+B+C+D	0.294	0.661	0.637	0.647	0.648
A+B+C+D+E	0.293	0.630	0.653	0.663	0.641
A+B+C+D+E+F	0.289	0.646	0.623	0.638	0.642
A+B+C+D+E+F+G	0.297	0.607	0.669	0.678	0.646

A: Evolutionary profiles; B: predicted relative accessible surface area; C: predicted secondary structure; C: physicochemical propensities; D: amino acid function composition; F: Sparse profile; G: Amino acid pair.