Bioinformatics Seminar
Time: 11AM
Venue: Zoom Webinar
30 November 2021
Analyzing epigenomic data with epigraHMM
Pedro BaldoniWEHI Bioinformatics
Epigenomic assays have been extensively used in the last decade to study interactions between DNA, proteins, and other cellular elements. Many of these interactions have been shown to regulate essential cellular functions and to be associated with complex diseases. Despite the long history of technological and methodological development in the area, the de novo analysis of certain epigenomic marks that exhibit broad regions of sequencing read enrichment (peaks) remains challenging. The analysis of such broad marks using assays such as ChIP-seq, DNase-seq and, most recently, CUT&Tag sequencing, is often complicated due to the low signal-to-noise ratio, local dependency of read enrichment, diversity of peak profiles, and non-linear biases. In this talk, I will present my recently developed Bioconductor package epigraHMM that tackles these challenges with a suite of efficient and flexible hidden Markov model peak callers to improve on current methods, while being robust for narrow and punctate peaks. epigraHMM makes use of the integration between R, C++ and HDF5 files to provide fast and light weight algorithms. We integrate seamlessly with other Bioconductor packages for an end-to-end analysis of epigenomic data under a variety of scenarios with tools for the detection of consensus and differential peaks, assessment of combinatorial patterns of differential enrichment across multiple conditions, and genomic segmentation.