Hi all,

I am currently trying to figure out how to use peaks that were identified by a first analysis for a second run as in the following example:

Imagine condition A being the control and condition B being a treatment for cells that can be countered via an inhibitor used in condition C. Now, to take into account possible off-target effects of the inhibitor, I want to select all peaks found as significantly altered in contrast A-B and test whether they show an opposing behavior in contrast B-C. Alternatively, I would also be fine with the intersect of both tests, however, in this case I would lose statistical power due to running more tests than necessary.

Thus far, I ran DiffBind on all pairwise contrasts as described in the manual, so I have results for A-B, A-C and B-C. The documentation on the mentioned issue seems a bit unclear to me, as I have tried to fiddle with the functions dba.overlap, dba.mask and dba.peakset yet none of them seems to be doing what I am looking for.

Any help is greatly appreciated!