Machine learning approach to predict protein phosphorylation sites by incorporating evolutionary information

BMC Bioinformatics

Table 3 Three fold cross validation performance of the prediction system using ratio 1:1.5.

Residue	W	Ac(%)	Sn(%)	Sp(%)	Mcc	FPR
	7	74.94	51.80	87.33	0.43	0.13
	9	75.02	52.17	87.26	0.44	0.13
S	11	75.54	53.00	87.61	0.45	0.12
	13	75.78	53.46	87.73	0.45	0.12
	15	76.15	54.53	87.74	0.46	0.12
	7	72.44	46.76	89.11	0.41	0.11
	9	72.47	48.14	88.26	0.41	0.12
T	11	72.39	50.35	86.70	0.41	0.13
	13	72.70	50.47	87.13	0.42	0.13
	15	72.63	50.12	87.24	0.41	0.13
	7	70.38	36.69	91.60	0.36	0.08
	9	71.76	41.45	90.84	0.39	0.09
Y	11	71.84	41.12	91.19	0.39	0.09
	13	72.31	42.11	91.33	0.40	0.09
	15	72.12	40.74	91.88	0.39	0.08

Here in this table, the sensitivity (Sn) and the specificity (Sp) columns reveal that the system using this ratio identifies most of the sites as negative.

ISSN: 1471-2105