I am using DiffBind to find the overlapped peaks.

I used the following commands:

mcf <- dba.peakset(mcf,peaks=peaks.v,peak.caller=peak.caller.v, sampID=sampID.v,tissue=tissue.v,

                          factor=factor.v,condition=condition.v,replicate=replicate.v)

dba.plotVenn(mcf,mask=c(17,19),main=paste0(colnames(mcf$class)[c(17,19)],collapse = “-vs-"))

After this, I got the following Figure

https://www.dropbox.com/s/i7e7xrjkp5hkmpm/Li_zhao.jpeg?dl=0

In my data,  cJun_Li has 283 peaks, GSE46166_Peaks_cJun_reff_input_200 has 15832 peaks originally.

After performing overlapping, there are only 7 peaks for  GSE46166_Peaks_cJun_reff_input_200 in this venn.

I am justing if there are something wrong with my command. 

I am justing wondering how the overlap between peaks are calcuated in DiffBind.

Thank you,

Aimin  

Hello Aimin-

I tracked this down to a bug in DiffBind that arises when a) dba.peakset() is used to add a specific set of peaks (rather than just the sample sheet being loaded in dba()) and b) the added peakset is not sorted.

I have already checked in a fix that will appear as DiffBind 2.4.8 in the next day or so to take care of the sorting.

Cheers-

Rory

Rory,

Yes, after I sort the peaks using DiffBind:::pv.peaksort.

It works perfectly.

Thank you very much.

Aimin