MISSEL: a method to identify a large number of small species-specific genomic subsequences and its application to viruses classification

Table 9 Classification accuracy on training and test set for the 5 genomic regions of polyomaviruses (mean ± standard deviation computed on all solutions)

Genomic region	β	Train [%]	Test [%]
LT	2	100	93.94
	4	100	92.93 ± 1.75
	6	100	93.18 ± 1.56
	8	100	93.94
	9	100	93.94
	11	100	93.94
	13	100	93.94
	15	100	93.94 ±1.50·10⁻¹⁴
	16	100	93.94
	18	100	93.94
	20	100	93.94 ±2.92·10⁻¹⁴
ST	3	100	93.75
	5	100	93.75
	8	100	93.75
	11	100	93.75
	12	100	93.75
	17	100	93.75
	20	100	93.75
VP1	2	100	90.32
	4	100	93.55
	6	100	93.55 ±2.97·10⁻¹⁴
	8	100	93.15 ± 1.45
	9	100	93.55
	11	100	92.38 ± 2.61
	12	100	93.55
	14	100	93.55
	17	100	93.55
	19	100	92.74 ± 2.20
	20	100	93.55
VP2	2	100	93.55
	4	100	93.55
	7	100	92.96 ± 1.31
	8	100	93.55
	10	100	93.55
	12	100	93.55
	13	100	93.55
	19	100	93.55
	20	100	93.55
VP3	2	100	93.33
	4	100	91.11 ± 1.92
	6	100	93.33
	8	100	93.33
	13	100	93.33
	15	100	93.33
	16	100	93.33
	17	100	93.33
	19	100	93.33

ISSN: 1756-0381