Assay

ChIP-seq: sanity checks and pitfalls

Antibody pull-down enriches DNA bound by a factor or carrying a histone mark; reads are aligned and peaks are called against an input or IgG control. Sharp peaks (transcription factors) and broad domains (H3K27me3, H3K36me3) need different caller settings, and the absence of a matched input turns copy-number variation and open-chromatin bias into fake peaks. Duplicate rate, fraction of reads in peaks and the cross-correlation profile tell you whether the immunoprecipitation worked before any biology is interpreted.

Who this is for: Transcription-factor and chromatin biologists asking where a factor binds or how a histone mark changes between conditions. Their datasets are old enough that genome builds and chromosome naming conventions get mixed, and the input control is the piece most often missing or mishandled.