Quantification And Comparison Of Chip-Seq Peaks

Question

[Ying W]

I have two chip-seq experiments under two conditions. I was wondering if it was possible to say that a peak is 'bigger' in one condition versus the other and if there is software that will quantify the difference. Current chip-seq peak callers seem to just say if there is a peak or not.

Thanks

Answer 1

[2184687-1231-83-]

One option is to use DESeq to analyse count data from the Chip-seq summits in your two conditions and test for differential intensity:

http://www-huber.embl.de/users/anders/DESeq/

[Ian]

@avilella +1 I have been using DESeq for ChIP-seq data (comparing read counts over two replicates of two time points [4 samples]). Apparently there is a GLM model that will allow the input control read counts to be factored in... Have you looked at this?

Answer 2

[Ian]

You could also try DIME or EdgeR.

I am currently trying DESeq, which seems to be producing sensible results. I am going to try DIME next as a comparison.

Answer 3

[Ying W]

Thanks, I will give DESeq a try,

I have found a paper that seems to address this issue but approaches it in an entirely different way

"Identifying dispersed epigenomic domains from ChIP-Seq data." http://www.ncbi.nlm.nih.gov/pubmed/21325299

Their software RSEG can be found here: http://smithlab.cmb.usc.edu/histone/rseg/

Answer 4

[Ziegfeld]

Also SICER. Rseg runs slower than SICER. You may want to use ChIPDiff which uses HMM and could be fed the output -island.bed files. You can tune confidence% and "fold change" parameters to solve your problem.

Answer 5

[dnaseiseq]

Bailey T, Krajewski P, Ladunga I, Lefebvre C, Li Q, et al. (2013) Practical Guidelines for the Comprehensive Analysis of ChIP-seq Data. PLoS Comput Biol 9(11): e1003326. doi:10.1371/journal.pcbi.1003326

http://www.ploscompbiol.org/article/info:doi/10.1371/journal.pcbi.1003326