Cscan: finding common regulators of a set of genes by using a collection of genome-wide ChIP-seq datasets

Abstract

The regulation of transcription of eukaryotic genes is a very complex process, which involves interactions between transcription factors (TFs) and DNA, as well as other epigenetic factors like histone modifications, DNA methylation, and so on, which nowadays can be studied and characterized with techniques like ChIP-Seq. Cscan is a web resource that includes a large collection of genome-wide ChIP-Seq experiments performed on TFs, histone modifications, RNA polymerases and others. Enriched peak regions from the ChIP-Seq experiments are crossed with the genomic coordinates of a set of input genes, to identify which of the experiments present a statistically significant number of peaks within the input genes' loci. The input can be a cluster of co-expressed genes, or any other set of genes sharing a common regulatory profile. Users can thus single out which TFs are likely to be common regulators of the genes, and their respective correlations. Also, by examining results on promoter activation, transcription, histone modifications, polymerase binding and so on, users can investigate the effect of the TFs (activation or repression of transcription) as well as of the cell or tissue specificity of the genes' regulation and expression. The web interface is free for use, and there is no login requirement. Available at: http://www.beaconlab.it/cscan.

Figure 1.

(a) Example of the output of Cscan showing the list of input genes (input box on the left hand side) as well as the TFs list (right, top) and ‘features’ (histone modifications, polymerase binding and so on, bottom right) ranked according to the resulting P value. See the main text for further explanation on the output fields. (b) The ‘Experiment view’ table, showing for a TF (E2F4) in the selected cell line (HeLa-S3) the target genes that were included in the input sample (left). The tables on the right show enrichment of other TFs (top) and features (bottom), computed on the E2F4 target genes.

Figure 2.

The most significantly enriched TFs in the different phases of human cell cycle (P < 10⁻⁵ in at least one set). We considered experiments performed on the same cell line of the expression data (HeLa). For TFs for which HeLa data are not yet available, we employed K562 data.