INSnet: a method for detecting insertions based on deep learning network

BMC Bioinformatics

Table 1 Performance comparison of SV callers on CLR dataset about HG002

Coverage	INSnet	cuteSV	SVIM	Sniffles
CLR
69X
Precision	0.9346	0.9247	0.9421	0.5849
Recalll	0.915	0.879	0.7967	0.8876
F1	0.9247	0.9012	0.8633	0.7051
35X
Precision	0.9095	0.9194	0.9204	0.5806
Recall	0.8644	0.8596	0.7961	0.8741
F1	0.8864	0.8762	0.8538	0.6977
20X
Precision	0.9084	0.8386	0.8366	0.5169
Recall	0.7838	0.8107	0.7547	0.8402
F1	0.8415	0.81	0.7936	0.64
10X
Precision	0.8822	0.9502	0.934	0.2906
Recall	0.6122	0.575	0.5223	0.7149
F1	0.7228	0.7062	0.6715	0.4132
5X
Precision	0.7643	0.6996	0.5988	0.4264
Recall	0.3943	0.4271	0.3911	0.4164
F1	0.5202	0.5201	0.4732	0.4213

ISSN: 1471-2105