Abstract:ChIP-seq technology provides an accurate characterization of transcription or epigenetic factors binding on genomic sequences. With integration of such ChIP-based and other high-throughput information, it would be dedicated to dissecting cross-interactions among multilevel regulators, genes and biological functions. Here, we devised an integrative web server CMGRN (constructing multilevel gene regulatory networks), to unravel hierarchical interactive networks at different regulatory levels. The newly developed method used the Bayesian network modeling to infer causal interrelationships among transcription factors or epigenetic modifications by using ChIP-seq data. Moreover, it used Bayesian hierarchical model with Gibbs sampling to incorporate binding signals of these regulators and gene expression profile together for reconstructing gene regulatory networks. The example applications indicate that CMGRN provides an effective web-based framework that is able to integrate heterogeneous high-throughput data and to reveal hierarchical 'regulome' and the associated gene expression programs.AVAILABILITY: http://bioinfo.icts.hkbu.edu.hk/cmgrn; http://www.byanbioinfo.org/cmgrn CONTACT: yanbinai6017@gmail.com or junwen@hku.hk Supplementary Information: Supplementary data are available at Bioinformatics online.

CMGRN: a web server for constructing multilevel gene regulatory networks using ChIP-seq and gene expression data