Comparing differentially expressed peaks to get unique values

Question

[rykerklie7]

I have done differential expression of the peaks from ATAC-seq data using DESeq2.

I have A vs B (Let us call this X), B vs C (Let us call this Y), and C vs D (Let us call this Z). I would like to compare X - (Y and Z) to get the unique peaks that are differentially expressed only in X.

I have the data in the following format

     Peak_name  Chr Start  End Width

I tried something as simple as the function merge in R but not sure if it is the right way to do it. And I am not sure if the overlap functions in diffbind works for this. I tried

      list1=merge(Y,Z, by.x = 0, by.y =0)
      list2=merge(X,list1, by.x=0, by.y=0)

Answer 1

[Rory Stark]

The overlap (dba.overlap()) and Venn diagram (dba.plotVenn()) functions in DiffBind can handle these types of queries, but in R it is probably simpler to turn your peaksets into GRanges objects and use the built-in overlapping functions.