histone-aggregationlisted

# Aggregate Histone ChIP-seq Peaks Across Studies ## When to Use - User wants to combine histone ChIP-seq peaks across multiple ENCODE experiments for a tissue or cell type - User asks "where is H3K27ac in pancreas?" or "build a histone mark map for liver" - User needs a union peak set from multiple donors, labs, or replicates - User wants to create a consensus binding map from multiple ChIP-seq datasets - Example queries: "aggregate all H3K4me3 peaks in brain", "combine histone marks across donors", "build enhancer map from H3K27ac data" Build a comprehensive map of histone mark binding for a tissue/cell type by merging narrowPeak files from multiple ENCODE experiments into a union peak set. ## Scientific Rationale **The question**: "Does my tissue have this histone mark, and at what genomic locations?" This is a **detection/cataloging** question, not a differential one. Once a histone mark passes noise thresholds (ENCODE IDR, quality metrics), detection is binary — the mark is either bound or not. If detected in one donor but not another, that region is still a real binding site. Individual variation and technical differences (lab, depth, antibody lot) explain *absence*, not that *presence* is spurious. **Therefore: we want the UNION of all detections, not a consensus.** ### Literature Support - **ChIP-Atlas** (Oki et al. 2018, EMBO Reports, 597 citations): Integrated >70,000 public ChIP-seq datasets using union of all peak calls - **ENCODE Phase 3** (Gorkin et al.