Skip to main content
Fig. 1 | BMC Genomics

Fig. 1

From: Mid-pass whole genome sequencing enables biomedical genetic studies of diverse populations

Fig. 1

Optimized methods for genotyping from mid-pass whole genome sequencing. a) Outline of strategy for variant calling and imputation using a combination of 100 high- and 1410 mid-pass sequenced genomes. Recall, precision, and non-reference concordance rate (NCR) calculated for imputed genotypes derived from mid-pass sequencing of the genomes of 92 individuals as a function of: pre-imputation call-level filtering using the GQ metric (keeping variants with GQ > X, b-d), binned sequencing coverage (e-g), and both coverage and with (MP + HP, dotted lines) or without (MP, solid lines) inclusion of high-pass sequencing from 100 individuals in joint-calling and imputation (h-j). For runs without high-pass included (MP), data for the 100 individuals was substituted with mid-pass data. Metrics were calculated using previously available genotype calls derived from 30x whole-genome sequencing as a truth set. All metrics plotted were calculated for SNVs only (for indels see Figs. S3, S5, and S6). For boxplots, bottom whisker: Q1–1.5*interquartile range (IQR), top whisker: Q3 + 1.5*IQR, box: IQR, center: median, and outliers are not plotted for ease of viewing

Back to article page