gene-level enrichment of H3K36Me3

[mrfox]

Hi guys,
I wonder if there is a software tool to measure the GENE-LEVEL enrichment of H3K36Me3 signals on CHIP-Seq data. I know MACS and SICER are good tools for peak calling but they are not specifically for gene-level signal enrichment measurement.

Any recommendation is appreciated.

[mudshark]

to the best of my knowledge there is no established and validated tool to perform this task.

however, one straight-forward approach would be to collect the sum the reads on all genes in IP and control sample and analyze that data with DESeq/edgeR.

maybe, you can even treat your raw data as if it were RNASeq data.

[Simon Anders]

Do you want to compare two conditions, or just score in a single sample?

If the latter: Why don't you just count how many ChIP reads you have falling onto each gene (and maybe divide by gene length and/or number of reads falling on to the same region in the input sample). What you then do with these score is another question, of course.

[mrfox]

Thank you for your reply mudshark and Simon. I want to compare two conditions. And I know DESeq is an option. But I am not sure exactly how to identify genes with differential H3K36Me3 signal between two conditions (I guess poisson test is more appropriate than negetive binomial for CHIP Seq).

So when counting the reads, a gene should be annotated like a single exon starting from the leftmost base pair of this gene and ending at the rightmost?

I have replicates for each condition. Simon, do you think I can perform the differential analysis like for RNA-Seq?

I appreciate your help.