Systematic benchmark of state-of-the-art variant calling pipelines identifies major factors affecting accuracy of coding sequence variant discovery

Table 1 Descriptive statistics of the gold standard sequencing datasets used in the study

Sample	Type	Source (SRA ID)	Mean coverage	Fraction of 10x bases	Median called variant count^b	True variant count
HG001	WGS	GIAB FTP	22.2	0.987	20,422	20,444
HG001	WES	ERR1905890	248.8^a	0.985^a	19,875	20,444
HG002	WGS	GIAB FTP	23.2	0.990	20,651	20,647
HG002	WES	SRR2962669	241.4	0.987	20,048	20,647
HG003	WGS	GIAB FTP	23.2	0.987	20,623	20,660
HG003	WES	SRR2962692	203.9	0.987	20,046	20,660
HG004	WGS	GIAB FTP	22.8	0.990	20,729	20,745
HG004	WES	SRR2962694	228.4	0.987	20,112	20,745
HG005	WGS	GIAB FTP	37.3	0.998	20,650	20,620
HG005	WES	SRR2962693	195.5	0.985	19,969	20,620
HG006	WGS	GIAB FTP	25.5	0.991	20,320	20,354
HG006	WES	SRR14724507	183.2	0.955	19,650	20,354
HG007	WGS	GIAB FTP	25.6	0.992	20,483	20,526
HG007	WES	SRR14724506	176	0.957	19,793	20,526

Coverage and variant statistics are given with respect to GIAB v4.2 high-confidence CDS regions. ^aCoverage values for downsampled HG001 WES dataset are given (see Methods); ^bvariant counts were obtained by calculating the median number of variants discovered by different pipelines. Full statistics for each sample and tool combination is given in Supplementary Table S1

ISSN: 1471-2164