Gene function prediction based on genomic context clustering and discriminative learning: an application to bacteriophages

Table 3 Prediction settings and results for the nine gene functions.

	Terminase	Portal	Head	Tail	Tape measure	Prohead protease	Lysin	Holin	Integrase
# positive samples	93	83	26	26	21	11	25	69	67
# negative samples	308	195	107	133	82	28	45	213	102
# clusters	17	15	7	6	7	4	6	16	12
Prediction Accuracy at t = 0.1(%)	86.9	85.89	67.87	83.33	75.68	66.67	100	79.5	82.18

The total number (#) of positive training samples, negative training samples and the number of clusters involved with each gene class are shown. Accuracy values are computed using leave-one-out cross validations. K-Means adaptive threshold t = 0.1. GRBF kernel's σ = 2 for Head and Tail; σ = 11.3 for all other cases.

ISSN: 1471-2105