Impact of index hopping and bias towards the reference allele on accuracy of genotype calls from low-coverage sequencing

Table 4 Concordance of best-guess genotype calls from sequence data with SNP array genotypes, using allele read counts obtained from aligned reads in BAM files

	n^a	Genotype concordance (%)	Allele concordance (%)	Concordance by genotype (%)
				True = 0			True = 1			True = 2
				0\|0	1\|0	2\|0	0\|1	1\|1	2\|1	0\|2	1\|2	2\|2
Low coverage
1×	28,300	62.1	80.8	99.34	–	0.66	51.46	–	48.54	0.96	–	99.04
2×	32,699	79.5	89.7	98.42	1.53	0.05	26.41	48.15	25.44	0.21	1.70	98.09
3×	25,993	88.3	94.1	98.25	1.72	0.03	14.01	71.98	14.01	0.12	2.36	97.52
4×	16,346	92.5	96.3	97.91	2.09	0.00	8.36	83.84	7.80	0.00	2.77	97.23
5×	8878	94.9	97.5	97.28	2.72	0.00	4.83	91.15	4.02	0.16	2.81	97.03
6–10×	6444	95.0	97.5	97.43	2.50	0.07	5.01	91.09	3.90	0.00	2.75	97.25
Overall	118,660	81.1	90.5	98.39	1.43	0.18	24.43	52.30	23.27	0.33	1.71	97.96
High coverage	131,782	99.8	99.9	99.80	0.19	0.01	0.12	99.81	0.07	0.10	0.17	99.73

^aNumber of genotypes called for 5531 SNPs across 26 individuals both for low- and high-coverage data
Concordance is shown by coverage at variant site

ISSN: 1297-9686