cross strand correlation in BrU-UV ChIP data

Question

[mforde84]

Hi,

We did a anti-BrU pulldown for two biological replicates, merged tag densities, and called peaks with input control from an IgG pulldown.

I generated a plot to assess strand cross correlation and shift. It appears that we have watson peaks at about 50bp which is consistent with the read length, and then we have crick peaks at ~150bp. This seems to suggest that the ChIP enrichment worked, however I'm somewhat confused by the jaggedness of the plot. I thought the curves were supposed to be alot smoother? Expecially since each biological replicate has 30-40M uniquely mapped reads.

- https://ibb.co/msVe6Q

Also, instead of using the IgG pulldown as the control, we have been discussing using the supernatant mRNA from the ChIP experiment as background. Essentially we BrU and UV cells, collect mRNA, and pulldown anti-BrU. So in theory the remaining supernatant should just be the remaining mRNA that are depleted for our BrU captured regions. In theory we can use this as an input control, yes? Or are we just begging the question by using a baseline for peak calling which we know is depleted for BrU?

edit:bump

[mforde84]

Really would appreciate some help with this. Thanks.