BreakNet: detecting deletions using long reads and a deep learning approach

BMC Bioinformatics

Table 2 Performance comparison of SV caller on HG002 data

	Coverage		BreakNet	SVIM	cuteSV	SNIFFLES
CLR	69X	Precision	0.9704	0.9678	0.9707	0.9604
		Recalll	0.9169	0.9341	0.9282	0.9224
		F1	0.9429	0.9507	0.9492	0.9410
	35X	Precision	0.9469	0.9653	0.9775	0.9556
		Recall	0.9169	0.9292	0.8955	0.9160
		F1	0.9316	0.9468	0.9351	0.9355
	20X	Precision	0.9524	0.9722	0.9790	0.9720
		Recall	0.8776	0.8389	0.8203	0.7983
		F1	0.9135	0.9004	0.8926	0.8770
	10X	Precision	0.9213	0.9790	0.9819	0.9785
		Recall	0.8134	0.6704	0.6646	0.6470
		F1	0.8640	0.7959	0.7925	0.7790
CCS	28X	Precision	0.9552	0.9400	0.9492	0.9020
		Recall	0.9350	0.9430	0.9336	0.8325
		F1	0.9450	0.9415	0.9414	0.8657
	10X	Precision	0.9424	0.9360	0.9609	0.9110
		Recall	0.9282	0.8940	0.8398	0.6357
		F1	0.9353	0.9146	0.8965	0.7490

ISSN: 1471-2105