We are now studying chromatin modification signature in normal vs disease human samples and found that DiffBind worked really well for our purpose.
Manual tells me that DiffBind uses the same statistics of EdgeR to find differential bound site, but I'm not sure how DiffBind treats those factors specific for ChIP-seq analysis.
1) Input reads
There seems to be some function like bScaleControl to manage them (is that correct?), but what is the basics of the calculation in which bamControl are treated?
2) peak scores
My peak files are in bed format which have peak scores in fifth column. How are they considered during DB calculation?
Please excuse me if they are already described somewhere but I'd appreciate if you could give some information. Thank you so much in advance.