BiG Talk – Best practices and common pitfalls in methylome sequencing analysis
November 10, 2025 @ 13:30 – 14:00 CET
Abstract
Whole-genome methylation sequencing provides single-base resolution of DNA methylation but remains one of the most computationally challenging omics data types to process. Its complexity stems from bisulfite-induced sequence conversion, protocol-specific artifacts, and the large number of available software tools, making reproducible analysis and cross-study integration particularly difficult. To address this we systematically benchmarked ten complete computational workflows across five distinct profiling protocols using an experimental gold-standard reference. The results reveal substantial performance differences in accuracy, computational efficiency, and downstream analysis, particularly for low-input samples. Furthermore, we introduced a continuous, “living” benchmarking platform for future software developments. To further facilitate large-scale, standardized methylome analysis, we developed complementary R package for efficient storage, manipulation, and integration of genome-wide methylation data that also supports applications such as methylation-based clocks, or reference-based cell-type deconvolution enabling tissue composition inference from whole-genome methylomes. Together, these resources address the difficulties of processing, analyzing, and interpreting DNA methylation data at genomic scale to ease reproducible epigenome research.
Biography
Reka Toth is a computational biologist and group leader at the Luxembourg Institute of Health (LIH). She heads the Multi-omics Data Analysis Group in the Department of Cancer Research and drives omics initiatives within the Medical Informatics, Bioinformatics & AI Unit. Trained as a molecular biologist, she completed a PhD in genetic epidemiology at the University of Debrecen (Hungary) focused on chronic diseases and addiction, and later added Software Engineering to pivot into computational biology.
In 2013, Dr. Toth joined the National Center for Tumor Diseases in Heidelberg, where she led the Biostatistics and Bioinformatics Group on colorectal cancer. She then moved to the Epigenomics Division at the German Cancer Research Center (DKFZ), expanding her work in computational epigenomics across cancer types, with a particular emphasis on DNA methylation for prognosis and cell-of-origin. She joined LIH in 2021 and established her own group in 2025. Her research focuses on methylation as a diagnostic and prognostic biomarker and on integrative multi-omics to reveal complex biomarker signatures and mechanisms in cancer.
Host: Louella Vasquez, NBIS (louella.vasquez@scilifelab.se)
Broadcast link (live event): https://lu-se.zoom.us/j/62767371752
The talk is sponsored by the SciLifeLab Epigenomics Symposium.
A recording will also be available on the NBIS YouTube channel.
info@nbis.se
Google Calendar 
