Meta-analysis of genome-wide chromatin data

Abstract

Genome-wide analyses of chromatin factor-binding sites or histone modifi cation localization generate lists of up to several thousand potential target genes. For many model organisms, large annotation databases are available to help with the characterization and classifi cation of genomic datasets. The term metaanalysis has been coined for this type of multi-database comparison. In this chapter, we describe a workfl ow to perform a transcriptional and functional analysis of genome-wide target genes. Sources of transcription data and clustering tools to subdivide genes according to their expression pattern are described. For a functional analysis, we focus on the Gene Ontology (GO) vocabulary and methods to uncover over- or underrepresented functions among target genes. Genomic targets of the histone modifi cation H3K27me3 are presented as a case study to demonstrate that meta-analysis can uncover functions that were hidden in genome-wide datasets.