Figure 3

Genotype mode performance. A) Comparison between htperc and binom options on WGS data. Comparison of number of heterozygous calls for htperc and binom (P = 0.01 and P = 0.05) methods on 7 WGS samples (numbers identify patients IDs) from [20] increasing the minimum depth of coverage (mdc). The inset shows the samples mean coverage computed on the original BAMs on the ~2.7 million SNPs of dbsnp 138 CEU. Labels 508(16X) and 508(8X) refer to samples data where reads were computationally down-sampled with probability equal to 0.5 and 0.25, respectively, from sample 508 (original mean coverage of ~33X). B) Estimation of sensitivity and False Discovery Rate (FDR) of ASEQ GENOTYPE mode. Each panel shows for individual sample the sensitivity and the FDR results of ASEQ GENOTYPE mode quantified for heterozygous calls obtained on WGS data with respect to the corresponding SNP array data calls by increasing the value of minimum depth of coverage. FDR curves for sample 508 (8X) are not shown as above the maximum considered FDR across the figure panels.