How super enhacer regions are detected

[Former user]

Hello ,

I am using ROSE to predict super enhancer region from peaks identified by HOMER of transcription factor ChIPseq data at two different time point. I uploaded the bedgraph, peak bed file and SE bed file. I have attached an image for one gene. I would like to understand why Super enhancer is not identified at both the time point for this gene instead of having exactly same number of peak near by this gene.

[Brian Abraham]

Hello,

Please view the methods of the corresponding paper that describes the method: http://www.cell.com/action/showMethods?pii=S0092-8674%2813%2900392-9

In brief, all enhancers in a dataset are ranked by their signal and the point at which the signal begins increasing rapidly is identified graphically. Differences between samples can affect the shape of this curve. Because super-enhancers are identified relative to the rest of the enhancers in a given sample, enhancers outside the region of interest can have a substantial effect on whether a region is considered a super-enhancer. For example, few exceptionally large enhancers can "push" enhancers below a previously established cutoff as the curve is "pulled" higher.