Analysis and prediction of antibacterial peptides

BMC Bioinformatics

Table 7 The performance of SVM, ANN and QM models using 15 residues of N+ C termini on independent dataset.

		SVM		QM		ANN
Dataset	Terminus	CPP*	PPV	CPP	PPV	CPP	PPV
39 original peptides of independent dataset	N	23	58.97	21	53.84	19	48.71
	C	30	76.92	29	74.35	27	69.23
	N+C	26	66.66	27	69.23	25	64.10
39 randomly extracted peptides from main dataset	N	35	89.74	33	84.61	34	87.17
	C	38	97.43	35	89.74	31	79.48
	N+C	36	92.30	34	87.17	36	92.30
24 peptides obtained from original independent dataset	N	21	87.50	20	83.33	17	70.83
	C	22	91.66	22	91.66	19	79.16
	N+C	22	91.66	22	91.66	22	91.66