Email updates

Keep up to date with the latest news and content from BMC Genomics and BioMed Central.

This article is part of the supplement: Eleventh International Conference on Bioinformatics (InCoB2012): Computational Biology

Open Access Proceedings

A novel unbiased measure for motif co-occurrence predicts combinatorial regulation of transcription

Alexis Vandenbon1*, Yutaro Kumagai23, Shizuo Akira23 and Daron M Standley1*

Author Affiliations

1 Laboratory of Systems Immunology, Immunology Frontier Research Center, Osaka University, 3-1 Yamada-oka, Suita, Osaka 565-0871, Japan

2 Laboratory of Host Defense, Immunology Frontier Research Center, Osaka University, 3-1 Yamada-oka, Suita, Osaka 565-0871, Japan

3 Department of Host Defense, Research Institute for Microbial Diseases, Osaka University, 3-1 Yamada-oka, Suita, Osaka 565-0871, Japan

For all author emails, please log on.

BMC Genomics 2012, 13(Suppl 7):S11  doi:10.1186/1471-2164-13-S7-S11

Published: 13 December 2012

Abstract

Background

Multiple transcription factors (TFs) are involved in the generation of gene expression patterns, such as tissue-specific gene expression and pleiotropic immune responses. However, how combinations of TFs orchestrate diverse gene expression patterns is poorly understood. Here we propose a new measure for regulatory motif co-occurrence and a new methodology to systematically identify TF pairs significantly co-occurring in a set of promoter sequences.

Results

Initial analyses suggest that non-CpG promoters have a higher potential for combinatorial regulation than CpG island-associated promoters, and that co-occurrences are strongly influenced by motif similarity. We applied our method to large-scale gene expression data from various tissues, and showed how our measure for motif co-occurrence is not biased by motif over-representation. Our method identified, amongst others, the binding motifs of HNF1 and FOXP1 to be significantly co-occurring in promoters of liver/kidney specific genes. Binding sites tend to be positioned proximally to each other, suggesting interactions exist between this pair of transcription factors. Moreover, the binding sites of several TFs were found to co-occur with NF-κB and IRF sites in sets of genes with similar expression patterns in dendritic cells after Toll-like receptor stimulation. Of these, we experimentally verified that CCAAT enhancer binding protein alpha positively regulates its target promoters synergistically with NF-κB.

Conclusions

Both computational and experimental results indicate that the proposed method can clarify TF interactions that could not be observed by currently available prediction methods.