ChIP-seq annotation, priority of the different genomic regions

Question

[xiangwangbio]

I am using Homer to perform peak annotation. I am little confused about the priority of the different genomic regions.

If one peak localizes at promoter for A gene and gene body for B, which annotation is more reasonable, promoter A or gene body B?

Answer 1

[Sinji]

I suppose it depends on your research question and what you suspect your antibody to do in your model system.

That being said, I believe that HOMER annotates peaks to their nearest TSS, thus the annotation would show the peak associated with gene A and not gene B.

[xiangwangbio]

Thanks a lot Sinji! My antibody has both promoter and gene body distribution. So maybe I should try different way. HOMER annotates peaks to their nearest TSS. Do you know any other package using different strategy?

[Sinji]

You may be able to check out the Bioconductor package ChIPseeker, it might be able to do what you want, though I'm not entirely sure.

The way I would do this is use something like bedtools closest and include more than 1 annotation per peak. So for instance you would want to include both gene A and gene B because your peak overlaps regions of both.

If bedtools doesn't do what you want, look into the bedops suite and then repost this question with the tag bedops and Alex should be able to help you solve your question using his bedops suite.

[xiangwangbio]

Hi Sinji, I tried ChIPseeker. It works well. I really appreciate your suggestion and help!