fold coverage chipseq

Question

[kanwarjag]

I ran a Chipsq experiment on several samples, when I look at the fold coverage of genome it turns out to be 1.4, 1.5 or 0.8 X I looked at one paper http://www.sciencedirect.com/science/article/pii/S0092867412012986 it has Supplementary Table S1 showing 0.39.... 1.01X coverage. I think it is a very low fraction of the genome which have been sequenced. Am I missing something in my interpretation. Any feedback/ guidance will be helpful.

Answer 1

[dariober]

It's more or less normal to have ~1x coverage genome wide. If you map around 30M reads, as per encode guidelines, you get 30M * 100bp / (3*1e9) ~ 1x coverage. Ideally, all these reads should map to peaks, which would give huge peaks and no background. In practice only ~5% of the reads map to enriched regions, which is not much but still ok if the pull down works reasonably well.