Theoretical characterisation of strand cross-correlation in ChIP-seq

摘要

With the development of next-generation sequencing (NGS) technologies and associated dramatic cost reductions, chromatin immunoprecipitation followed by sequencing (ChIP-seq) has become a major method to obtain genome-wide profiles of DNA-binding elements. Although ChIP-seq application studies have revealed massive biological insights, its experimental complexity and low signal-to-noise ratio (S/N), as compared with other NGS applications, still make ChIP-seq analyses difficult [1, 2]. Furthermore, differences among experimental protocols and subsequent data analyses could extensively bias the final peak calling results [3, 4]. As the sample size per study has increased, as in the cases of ENCODE [5] and ROADMAP [6], the need for a quality control (QC) methodology for ChIP-seq has become more important [7].