Pol_II genome wide occupancy analysis

Question

I would like to know how to perform this global Pol II occupancy which is shown here:

Paper

Figure 1G

So in this figure they are analysing the Pol II occupancy in Control,PAF1 KD and LEO1.

I have similar data.

In order to determine whether PAF1 has a direct effect on the regulation of Pol II at the promoter-proximal region, ChIP-seq was performed. Heatmaps ranked by decreasing Pol II occupancy show a positive correlation with PAF1 and Pol II (Figure 1G). To corroborate these findings, we also performed ChIP-seq with LEO1, as this was the only other antibody that we tested that gave any significant ChIP-seq signal. Like PAF1, LEO1 is found throughout gene bodies with particular enrichment around the TSS (Figure 1G). Interestingly, LEO1 was not broadly lost from chromatin upon PAF1 knockdown, but the distribution was altered, including reduced occupancy in the promoter-proximal region

So what I have done so far in terms of analysis is I did peak-calling using homer for my control and KD condition along with their Pol II chip.I was going for this differential binding analysis but this is now how i Pol II data is done after reading the paper I got. In homer there is this function analyzeRepeats.pl which is provided in this section Measuring Gene Expression in Exons vs. Gene Bodies. telling genes (default) - Counts tags on the full gene body (TSS to TTS).This is useful for GRO-Seq where we expect coverage across the entire transcript.Can also be used to quantify H3K36me3 or PolII ChIP-Seq.

How do I proceed and what tool should I use to show the Pol II occupancy at different condition and as of now I don't have respective rna seq data, do i need expression data of genes to perform occupancy analysis.

Answer 1

comment discussion between @krushnach-chandra and @stupidwolf (in chronical order):

SW: No you don't need gene expression data.. You can just count the reads from polII chip (normalized by some method of your choice) and perform the analysis

KC: are you suggesting I can compare the Pol II chip between control and Knock out with their respect IP data?

SW: more or less yes...if you are interested in genomewide effects.. For example, if the treatment causes the polymerase to stall, you will see more reads in the promoter compared to the gene body

KC: so naive question do i need those Input and IgG at all for the comparison?

SW: good question.. depends on how you analyze it. If you always normalize within the gene, for example, you calculate a coverage normalized to that in the gene, or some meta gene thing, you don't need.

SW: If you do a differential gene expression based analysis, in theory you don't really need that, because you will always compare between your two conditions right. The IgG is good for checking you don't have any odd stuff... You can see whether there are genes where the control is enriched over your samples

KC: "calculate a coverage normalized to that in the gene, or some meta gene thing" any example can you cite me..so that i can follow and proceed ,as it looks like binding data are really complicated compared to rna seq

KC: "If you do a differential gene expression based analysis," you saying i can take out counts and do diff analysis for chip seq data using deseq2? but how do i get the information regarding if Pol II activity is going or stalled ? and how do i find the occupancy from the differential result

SW: welcome to the world of epigenomics!!! maybe something like this

SW: As for differential gene expression, yes. If you do this separately for promoters and gene body, you can see that the fold change will be different for a gene's promoter vs gene body, in situations where there's stalling, or reduced elongation etc.. I think it's a good way to account for the variation

KC: your insights are good start for me to run the analysis

KC: mundane question how do i get promoter and gene body information is it after annotation which should be done after running analysis ?

SW: Meaning? I normally align the reads, then for TSS and gene body you have to define it yourself using the transcript or gene annotation. Normally like +/- 200 bp around TSS for promoter and gene body is rest of it?

KC: "Normally like +/- 200 bp around TSS for promoter and gene body is rest of it" so i can do that in both chipseeker as well as homer?