Automatic annotation of protein motif function with Gene Ontology terms

BMC Bioinformatics

Table 4 Top 5 GO terms associated with the motif PS000383 ranked according the conditional probability of correctness of association. The column 2~7 consist of the feature vector for motif-GO association is listed, the conditional probability p(Y = 1|X) is calculated with trained model and the true classes are list in right two columns of the table. The definition of the GO terms is listed at the bottom of the table.

GO Terms	Input Features (X)						p(Y = 1\|X)	True Class
	NT-M	NT	NM	NG\|M	NP\|T	M.I
GO:0006470	65	122	191	95	27	0.007268	0.97577	1
GO:0005001	23	24	191	95	5	0.003177	0.87681	1
GO:0004726	10	11	191	95	10	0.001331	0.68682	1
GO:0005634	11	1785	191	95	281	5.08E-06	0.27882	0
GO:0005887	17	1162	191	95	225	0.000193	0.15536	0

GO:0006470: Protein amino acid dephosphorylation; Cellular Processes
GO:0005001: transmembrane receptor protein tyrosine phosphatase activity: Molecular Function
GO:0004726: non-membrane spanning protein tyrosine phosphatase activity: Molecular Function
GO:0005634: Nucleus; Cellular Component
GO:0005887: Integral to plasma memberane; Cellular Component

ISSN: 1471-2105