Improvement of the ChIP-seq technique for histone posttranslational modifications through enhanced laboratory protocols and data processing methods

Massively parallel sequencing has rapidly become the dominant technique in various omics studies as it provides an unequalled amount of high quality quantitative data for a reasonable cost that is decreasing ever since the first next generation sequencers appeared. Subsequently, chromatin immunoprecipitation coupled with massively parallel sequencing (ChIP-seq) is the prevailing method of choice for investigating protein–DNA interactions in a genome-wide manner. Bioinformatics tools are also evolving quickly to meet the increasing demands of processing huge amounts of ChIP-seq data and to open the way for novel techniques and insights. However areas still exist that could benefit from improved wet-lab and dry-lab methods. One such area is data visualisation and interpretation; another is the ChIP-seq study of histone posttranslational modifications, especially the research of inactive histone marks which tend to produce diffuse broad enrichments instead of point-source peaks. Achieving proper enrichment and unbiased analysis in such histone mark studies proves to be a great challenge.
In this doctoral thesis we show how we addressed these issues on both the level of bioinformatics and the level of sample processing methods. We present our innovative analysis tools we developed to this end, among others a highly customisable, feature rich viewer for next generation sequencing data visualisation, and an analysis pipeline specifically aimed to handle broad enrichments from ChIP-seq studies of (inactive) histone marks. We propose specific software and specific peak calling settings to detect a range of histone modifications accurately, and we describe the way to determine the optimal settings. Along the pipeline we also present a protocol designed to enhance enrichments and facilitate peak detection in broad peak studies typical of inactive histone marks. We demonstrate how this method affects various enrichment types and propose potential applications that could benefit from it. Furthermore we show diverse achievements with the analysis pipeline, including the interpretation of the aforementioned wet-lab method, and the development of an automated ChIP-seq protocol optimised for low cell numbers.