MEME-ChIP: motif analysis of large DNA datasets

MEME-ChIP: motif analysis of large DNA datasets
Philip Machanick1 and Timothy L. Bailey1,*
+ Author Affiliations

1Institute for Molecular Bioscience, The University of Queensland, Brisbane 4072, Queensland, Australia.
↵*To whom correspondence should be addressed. Dr. Timothy Bailey, E-mail: t.bailey@imb.uq.edu.au
Received January 23, 2011.
Revision received March 18, 2011.
Accepted April 6, 2011.
Abstract

Motivation: Advances in high-throughput sequencing have resulted in rapid growth in large, high-quality datasets including those arising from transcription factor (TF) ChIP-seq experiments. While there are many existing tools for discovering TF binding site motifs in such datasets, most web-based tools cannot directly process such large datasets.

Results: The MEME-ChIP webservice is designed to analyse ChIP-seq “peak regions”—short genomic regions surrounding declared ChIP-seq “peaks”. Given a set of genomic regions, it performs 1) ab initio motif discovery, 2) motif enrichment analysis, 3) motif visualization, 4) binding affinity analysis and 5) motif identification. It runs two complementary motif discovery algorithms on the input data—MEME and DREME—and uses the motifs they discover in subsequent visualization, binding affinity and identification steps. MEME-ChIP also performs motif enrichment analysis using the AME algorithm, which can detect very low levels of enrichment of binding sites for TFs with known DNA-binding motifs. Importantly, unlike with the MEME webservice, there is no restriction on the size or number of uploaded sequences, allowing very large ChIP-seq datasets to be analyzed. The analyses performed by MEME-ChIP provide the user with a varied view of the binding and regulatory activity of the ChIP-ed TF, as well as the possible involvement of other DNA-binding TFs.

Availability: MEME-ChIP is available as part of the MEME Suite at http://meme.nbcr.net.

Contact: t.bailey@uq.edu.au