Ultra-deep mutant spectrum profiling: improving sequencing accuracy using overlapping read pairs

BMC Genomics

Table 4 Summary of coverage, total number of candidate variants (all polymorphisms) and number variants called by the variant detection model in Q30 ORP for the 3 natural viral samples

Natural sample	Mean ORP coverage	# candidate variant calls at Q≥30	# variants called at error rate 5×10^-5	# variants called at error rate 5×10^-4	FDR for error rate 5×10^-5	FDR for error rate 5×10^-4
BCV	79,037x	20413	2133	152	19%	0%
Fox1	80,546x	14472	1354	70	21%	0%
Fox2	96,198x	20417	1596	88	26%	0%

The number of candidate variants served as the Bonferroni correction factors in the variant detection model. The number of false discoveries in each sample at error rate 5×10^-5 is estimated by taking 2% (coverage adjusted false positive rate derived from control plasmids, Additional file 1: Table S2, ) of the number of candidate variants. The false discovery rate (FDR) is the proportion of false discoveries among variants discovered, e.g. for BCV, it’s the ratio between 2% of 20413 and 2133, i.e.19%. FDR at 5×10^-4 is 0% for all samples since the estimated false positive rate at this error rate is 0% (Additional file 1: Table S2).

ISSN: 1471-2164