I am creating a coverage plot +/- 250bp around a transcription start site with various ATAC-seq replicates. I have created these plots very easily for S2 (drosophila embryonic) cells using microarray data. However, I am having a hard time locating precise peak centers in the ATAC data which is very noisy. Ideally, I would like a smooth curve centered at 0bp. Right now, my plots seem to be really skewed right or left depending on the replicate (I am actually finding the peak centers by hand, using a BW visualization file which is clearly not working nor is that efficient) Is there any better, consistent, way of doing this? I am calling peaks using HOMER and creating the coverage plot using perl and python scripts.
If you call peaks using macs2, it also give the summit (
--call-summit
) , which is the peak center.