Combined Tss Plots For Chip Seq Peaks

Question

[aditi.qamra]

Hi,

I need to draw a TSS plot for my chip seq peaks, however the problem that im facing is -

I used CCAT peak calls on 10 samples and have shortlisted a list of 50 peaks that Im interested in. These 50 peaks are not necessarily being called in all 10 samples. Also I took a union of overlapping CCAT peaks across 10 samples ( ie peaks are a union of separate peaks across <= 10 samples depending on whether they have been called in a sample or not) to get a list of peaks from which I narrowed down onto 50.

So now that I want a TSS plot - I was considering using CEAS for it which needs a wiggle file. I right now can generate wiggle files for each sample ie. 10 wiggle files ( each one including all 50 peaks) BUT I wanted to get a common wiggle file for these set of 50 peaks. I mean my objective of having one single TSS plot ( instead of 10) for my 50 peaks is to get a general sense of the distribution of my peaks around TSS

What I was considering doing was taking the FC values from the CCAT peak calls and getting an average FC of the peaks that were combined to form an union peak and then use the average FC as a quantitative signal in wiggle files. So essentially add that information to my bed file (chr start stop avgFC ) and convert it to a bedgraph and further a wig file.

Does that seem like a right approach ?

Answer 1

[Ryan Dale]

One option is to modify this answer to a similar question. This will work with raw reads in BAM format or bigWig.

You could calculate one array for each sample using only the peaks called in that sample, concatenate all the arrays together, and then take the average across all of them to get the kind of average plot that CEAS provides.