Poster #31: Thomas Roule PhD

GNOMES: A Pipeline for Normalization and Differential Binding Analysis of CUT&RUN and ChIP-seq Data

Thomas Roule1 and Naiara Akizu12 1Perelman Center for Cellular and Molecular Therapeutics and Center for Brain Research in Development, Genetics and Engineering, Children’s Hospital of Philadelphia, Philadelphia, PA 19104, USA 2Department of Pathology and Laboratory Medicine and Genetics and Epigenetics Graduate Group, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA 19104, USA

Quantitative comparison of epigenomic datasets such as ChIP-seq and CUT&RUN remains challenging because sequencing-depth normalization often fails to account for differences in signal-to-noise ratios across samples and conditions. Spike-in controls can help, but are not always available and add experimental and computational complexity. In addition, normalization and differential binding are often performed with separate tools, limiting users’ ability to visually inspect normalized signal tracks and assess how normalization affects downstream results. Results: We developed GNOMES, Genome-wide NOrmalization of Mapped Epigenomic Signals, an integrated framework for normalization, quality control, visualization, and differential binding analysis of ChIP-seq and CUT&RUN data. GNOMES processes aligned reads and generates normalized coverage tracks, median signal tracks, candidate regions, and differential binding results. Its normalization strategy is based on percentile scaling of genome-wide local signal maxima, providing robust normalization across biological replicates and conditions. GNOMES supports single- and paired-end data, does not require input or IgG controls, and can be applied to both broad histone marks and narrow transcription factor profiles. The workflow includes optional consensus peak identification and integrates established tools such as MACS2 for candidate region detection and DESeq2 or edgeR for statistical testing. At each step, GNOMES produces quality-control outputs, including normalized bigWig files, BED files of significantly altered regions, heatmaps, PCA plots, and diagnostic summaries. The framework is highly configurable, organism-agnostic, and applicable to epigenomic datasets from diverse model systems. Conclusions: GNOMES provides a reproducible and customizable workflow for quantitative normalization, visualization, and differential binding analysis of CUT&RUN and ChIP-seq data, facilitating the identification and interpretation of chromatin changes across conditions.