Input giving peak at TSS! Yikes!

[pshah]

Hi all,
It's my first attempt at ChIP-seq. I optimized fixation time, shearing, IP. I confirmed my ChIP was working by qPCR. Prepped libraries and sequenced on a HiSeq 1500 (SE, 50 cycles to start just to make sure library prep gives me results I expect).

I analyzed the resulting data as follows:
1. map reads using bowtie (identical sequences not yet collapse, working on it!)
2. convert to bam format
3. plot profiles using ngsplot

I got some unexcepted results for my input, which look very similar to my ChIP samples. The major problem is that I am getting a peak at TSS and smaller peak at TES. I was expecting the input profile to be flat.

Is this a common problem? Do I have contamination in my library prep?

Here are my profiles of input and PAF1 ChIP (transcription elongation factor responsible for H3K4me3).

Thanks in advance for any wisdom you have to share.

[Chipper]

Yes, it is common. Just look at some ENCODE datasets. Sort of like FAIRE. You should expect your ChIP peaks to be much higher though. Would guess PAF1 has a few good sites at ~- 500 bp that averages out to similar enrichment as the TSS?

[pshah]

Thanks for your input, Chipper. It's good to know that this is a fairly common problem.

I do get some nice peak calling using MACS2. I will have to take a closer look at those peaks to see if I am really getting enough enrichment.