Comparative study on ChIP-seq data: normalization and binding pattern characterization

• DOI: 10.1093/bioinformatics/btp384
• Publication Date: 2009-06-27T00:24:20Z
• Authors: Cenny Taslim; Jiejun Wu; Pearlly Yan; Greg Singer; Jeffrey Parvin; Tim Huang; Shili Lin; Kun Huang
• Keywords: Normalization

Abstract Motivation: Antibody-based Chromatin Immunoprecipitation assay followed by high-throughput sequencing technology (ChIP-seq) is a relatively new method to study the binding patterns of specific protein molecules over entire genome. ChIP-seq allows scientist get more comprehensive results in shorter time. Here, we present non-linear normalization algorithm and mixture modeling for comparing data from multiple samples characterizing genes based on their RNA polymerase II (Pol II) patterns. Results: We apply two-step locally weighted regression (LOESS) approach compare across model difference using an Exponential-NormalK model. Fitted used identify associated with differential sites local false discovery rate (fdr). These are then standardized hierarchically clustered characterize Pol As case study, analysis procedure normal breast cancer (MCF7) tamoxifen-resistant (OHT) cell line. find enriched regions that (P < 0.0001). Our findings also imply there may be dysregulation cycle gene expression control pathways cells. show can analyze samples. Availability: Data available at http://www.bmi.osu.edu/~khuang/Data/ChIP/RNAPII/ Contact: [e-mail]; [e-mail] Supplementary information: Bioinformatics online.