Efficient and flexible Integration of variant characteristics in rare variant association studies using integrated nested Laplace approximation
Fig 1
rvGWAS workflow and QC plots for 1810 high quality samples from 1000GP used for benchmarking.
The rvGWAS workflow is exemplary shown for a simulated breast cancer case-control cohort based on 1810 whole exome sequencing datasets from 1000GP. (A) rvGWAS workflow for performing QC and six RVAS tests. Different colors indicate that cases carry different potentially damaging rare variants within the same functional biological unit, such as a gene. The QC module computes quality statistics shown in panels B-F. The result of each RVAS test is a ranked list of genes with various informative attributes. (B) Bar-plot for number of variants per sample, colored by functional annotation of variants. (C) Barplot for number of variants per sample, colored by assignment to cases (~1/2) or controls (~1/2). (D) Number of variants per gene in cases (x-axis) and controls (y-axis). Each dot is one gene, while the red line shows the ratio of the number of cases and controls (1:1). (E) Histogram for number of mutations per sample after removal of outliers. (F) Projection on first 10 PCA components. Samples are colored by sequencing center. The graph in the upper right corner shows the cumulative percentage of variance explained per principal component. Principal components can be used as covariates in several RVAS tests.