eclip-seq: Using edgeR/DESeq2 for normalizing peaks of IP/Input of transcript level

Question

[gnilihzeux]

Hello everyone, I've been unprofessional with data for some time. There's a problem stopped me of how to normalized the peaks of IP/input in eclip-seq.
I have a little experience in ChIP-seq, which normalized the peaks of IP/input using poisson's distribution on the level of genome region level, such as MACS2 with 1kb or more.
For eclip, what of my thinking is that it's similar to RNA-seq in gene or transcirpt level or other genome element such as exon. Peaks of transcript or gene in eclip-seq is actually similar to expression or counts of RNA-seq.
That's just my hypothesis but I hadn't find the same materials from literature or google.
Cound I ?
Thank you all.

Answer 1

[gnilihzeux]

After a few days study theories of edgeR/DESeq2, I think it's okay to test the difference of IP/Input of genomic tags.
However there is still a problem that the statistics power is uable to evaluate without replicates.
So, it's hard for me now to determine real peaks.
Still on the way...

[gnilihzeux]

It seems be okay to do so validated by viewing the peaks directly by IGV. Still there may some false positive result exsits, but I think it's okay.