ChIPseq - Identical peaks in IP and input samples

Question

[dorota.komar]

Hi guys, I have obtained some strange results from my ChIP-seq analysis: the ChIP sample looks exactly the same as Input sample, bacisally covering uniformly the entire genome. I have already performed other ChIP-seq experiment with the same line and antibody - that time it worked nicely (but we got some unexpected results, so wanted to make sure they are not an artefact, but a real protein behaviour). I have also checked the immunoprecipited DNA by q-PCR and could see clear difference in DNA abundance of positive and negative regions.

We sent 4 samples for sequencing: 2 ChIP and 2 inputs. The library preparation of one of ChIP samples did not work, so we have finally only sequenced 1 ChIP and 1 input. The very simple explanation could be that the sequencing service made a mistake and sequenced both inputs, or that I mislabelled the samples, but since one momber of the lab got the same results some time in the past, I am thinking there might be more difficult answer to that.

I was thinking since the binding of my protein is pretty low, the peaks were not wery high and with the high background, after amplification during library preparation and very deep sequencing it could be the expected failed results. But maybe there is another explanation that could save me from that in teh future. Any thoughts?

[Arup Ghosh]

So, you have just aligned the sequenced files and visualized using a browser like IGV?

Answer 1

[Devon Ryan]

The likely explanations are:

1. Mislabelled samples
2. Poor IP (have a look at this with plotFingerprint in deepTools).

There's no good way to differentiate post hoc whether you have a poor IP or the signal was lost during PCR. I mean, you could look at your duplication levels. If they're high then presumably you over amplified and perhaps that's causing the problem. It's also usually useful to look at a control region in your samples (whatever you're using for your qPCR).