[ChIP-seq. signal plot] How to merge ChIP-seq. biological replicates for signal profile plotting?

Question

[chiefcat]

Dear all,

In ChIP-seq. analysis, what is the proper way to combine biological replicates for signal profile plotting?

At the moment, I'm using deepTools plotProfile to visualize the signal density around TSS. After plotting the replicates separately, I can see that they are highly similar. Then, I would like to pool the replicates and presenting them as one single line in the profile. Should I achieve this by merging the bigWig files (e.g. bigWigMerge from UCSC tools as suggested by Devon in this post?) generated from bamCompare? (Each of the replicates has been compared to their own input in bamCompare & normalized to 1x Reads Per Genomic Content (RPGC).

Thanks!

Answer 1

[Devon Ryan]

I'd normally suggest keeping them separate. While you can merge them (and then scale accordingly, since bigWigMerge gives you an easy-to-manipulate text file), I'm not of the opinion that you're actually gaining much by doing so. Keeping samples separate allows people to easily judge biological variability, which is what they really want to get out of figures.

[chiefcat]

Hi Devon, thanks for your reply. Should I scale them according to no. of reads?

[Devon Ryan]

They're already 1x normalized, so merge and divide by the number of replicates.