I have a number of RNAseq bigwigs and ChIPseq bigwigs and would like to compare overall how counts in one location compare to counts in the same location in the contrasting experiment. To plot literally something like the figure here :
RPKMs you can get by transforming the counts from featureCounts. You can also use featureCounts to compute counts for the ChIPseq samples. For some ChIP signals they seem to just use log2(counts), for others counts per base, it's unclear if they just mucked up the labels or if they actually do change between the two. Either way you're just transforming counts.
Would you do any normalization to the ChIP? Or just take raw counts?
It won't really matter if you normalize things or not, that'll just shift the scale on the axes a bit while keeping the overall picture the same.