A De Novo Genome Assembly Algorithm for Repeats and Nonrepeats

Table 4

The effect of sliding window to SWA.

Window size	Repeat			Contigs (kb)		Accuracy (%)			Genome coverage (%)
	TRC	CNRC	NNC	N50	Max	CN-accuracy	Rep-accuracy	C-accuracy
3	11	Appendix	1016	0.4	3.0	100	76.1	100	81.2
7	10	Appendix	688	1	4.0	100	98.7	100	101
11	15	Appendix	612	1.1	4.0	100	89.6	100	101

Sequence length 𝐿=500 kb containing four kinds of repeats. Size of sliding window varies from 3 to 11. Contigs of repeat and nonrepeat are generated in an independent way by SWA at different levels with basic parameters: sequencing depth 𝑆_𝑑=0.5, filtered times = 1, read length 𝐿_𝑟=60, and 𝑘-mer=20. Contigs smaller than 200 are removed.