Software

I write R packages for analyses that I run often, so that other people can repeat them. All of them are open source.

ggwas

ggwas draws plots for genome-wide association studies with ggplot2. It covers the usual steps from quality control to figures for a paper: Manhattan, QQ, Miami and locus zoom plots, PheWAS, fine-mapping, colocalization and genetic correlations. It includes journal themes and colour-blind-safe palettes, and it downsamples points so that biobank-sized files still plot quickly. The package is under review at Bioconductor.

Code Documentation Zenodo DOI

ggwas in a few lines

pak::pak("bczech/ggwas")

library(ggwas)
data(example_gwas)  # columns CHR, BP, SNP, P

manhattan_plot(example_gwas)
qq_plot(example_gwas, show_lambda = TRUE)

Manhattan plot of the bundled example data, from the ggwas documentation.

Gallery

Manhattan plot

Miami plot: two traits

QQ plot

PheWAS

Genetic correlations

Circular Manhattan plot

BRASSVis

BRASSVis plots gene fusions and structural variants found by BRASS in whole-genome sequencing data. Each fusion gets one page with chromosome ideograms, exon structure, a circos plot, retained protein domains and read support. The layout follows the Arriba fusion plots, adapted to BRASS output.

Code

gDR

I contribute to gDR, an open-source R suite for processing and analysing drug response screens, available on Bioconductor. It processed the screens behind the papers in Cancers (2024) and Nature Communications (2025).

gdrplatform on GitHub

Internal tools

In pharmaceutical R&D I build R/Shiny applications for scientists and maintain production WDL pipelines for NGS data.