commonPeak: Equivalence testing to identify common ChIP-seq peaks across conditions and protocols

Abstract

BackgroundNew ChIP-seq protocols are increasingly adopted alongside established workflows. However, dedicated methods are lacking to quantify the agreement of peak location and intensity across datasets, including comparisons across protocols and biological conditions. ObjectivesWe present commonPeak, a statistical framework that identifies shared peaks across samples and tests whether their enrichment is similar across conditions, thereby supporting bench-marking of ChIP-seq protocols and cross-condition comparisons. ResultscommonPeak operates on BED peak sets and BAM files. As a use case, we applied it to an estrogen receptor alpha (ER) ChIP-seq dataset from tamoxifen-sensitive and -resistant breast cancer cell line MCF-7 cells and identified 225 shared peaks with significantly similar enrichment. These peaks were largely distinct from differentially bound sites and enriched for estrogen signaling. This illustrates how equivalence-based peak selection can help separate conserved ER-driven programs from condition-specific changes in ER-targeting endocrine drugs such as tamoxifen. Availability and implementationcommonPeak is freely available for non-commercial use via GitHub, with documentation and usage examples.