De-novo motif analysis was carried out on the central 100bp peak regions using the Homer software. For any ChIP-seq experiment in HAEMCODE, all peak regions were used as input to the findMotifsGenome.pl program in Homer. The program then looks for consensus sequences that are enriched in the sample relative to background (sequences randomly selected from the genome). The HTML page of the results and motif position weight matrices (PWM) is displayed on HAEMCODE.

For more information about Homer and results, see this page:
http://biowhat.ucsd.edu/homer/motif/index.html.

Although Homer performs searches for known motifs, HAEMCODE also displays results from an independent search conducted on a larger set of known motifs using TOMTOM. Known motifs were collected in-house from publications and public resources - Jaspar, UniProbe and Jolma et al.. PWMs of enriched motifs found by Homer were converted to MEME format and used as input to TOMTOM. Significant similarity of enriched motifs to known motifs are reported in a HTML output on HAEMCODE. Explanation of TOMTOM results can be found at the bottom of every results page.

For more information about TOMTOM, see this page:
http://meme.nbcr.net/meme/tomtom-intro.html.