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

BMC Bioinformatics

Table 5 Three fold cross validation performance of the prediction system using ratio 1:0.5.

Residue	W	Ac(%)	Sn(%)	Sp(%)	Mcc	FPR
	7	75.41	90.54	45.69	0.42	0.54
	9	75.61	89.35	48.62	0.43	0.51
S	11	76.45	89.65	50.52	0.45	0.49
	13	76.91	89.45	52.30	0.46	0.48
	15	77.21	90.03	52.04	0.47	0.48
	7	71.08	93.63	26.26	0.28	0.74
	9	70.89	92.40	28.14	0.28	0.72
T	11	71.58	93.24	28.54	0.30	0.71
	13	71.47	93.08	28.54	0.29	0.71
	15	71.45	93.59	27.44	0.29	0.73
	7	68.84	95.63	16.06	0.20	0.84
	9	70.58	94.26	23.93	0.27	0.76
Y	11	69.86	94.53	21.24	0.24	0.79
	13	69.57	94.21	21.02	0.23	0.79
	15	69.57	94.64	20.16	0.23	0.80

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 positive.

ISSN: 1471-2105