Analysis and visualization of ChIP-seq and ATAC-seq data using epiwraps

Author(s): Pierre-Luc Germain,Mark Robinson

Affiliation(s): ETH and University of Zürich, Switzerland



Workflows for epigenomics data, especially ATAC/ChIP-like data, typically involve a (sometimes clunky) mix of tools, within and outside R. This can create consistency or reproducibility issues, difficulties when trying to combine elements of different workflows, and complicates teaching. Although excellent R/Bioconductor-based solutions are available for many steps, some critical steps lack good R-based alternatives, and their integration often lacks the smoothness that Bioconductor has allowed us to enjoy in other subfields. Prompted by teaching bioinformatics and regulatory genomics, we developed a package (epiwraps, https://ethz-ins.github.io/epiwraps/ ) that aims to offer a set of flexible, user-friendly, and interoperable tools for some of the most basic functionalities. The package is not yet in Bioconductor, but has been in use for some years, and submission is expected well before the conference. The demo will showcase some of the main functionalities using ChIP-seq and ATAC-seq examples, while at the same time providing (a very condensed version of) some key lessons that we like to emphasize in our teaching. In particular, we will focus on 1) summarizing genomic signals, e.g. bigwig file generation; use them to 2) disentangle and interpret the different signals contained in ATAC-seq data; 3) discuss normalization methods, including bias-correction techniques (emphasizing the use of consistent normalization across tasks); and 4) visualization and clustering of genomic regions based on multiple signals.