Peak summits after IDR analysis

227 views
Skip to first unread message

Lucy

unread,
Sep 12, 2018, 12:28:58 PM9/12/18
to idr-discuss
Hi,

I have performed ATAC-seq on three different cell subsets and have generated an optimal peak list for each using IDR.  I then ran bedtools merge to generate a global peak list for all of the subsets.  However, by merging the peaks, I lose the summit information e.g. some peaks have multiple summits, but this information is lost when I run bedtools merge as these peaks are merged together.  Does anyone have any ideas about how to overcome this?  I would like to obtain the peak summits in order to perform motif analysis.

Many thanks,
Lucy

Anshul Kundaje

unread,
Sep 12, 2018, 12:45:22 PM9/12/18
to idr-d...@googlegroups.com
Take your merged peak list and bedtools intersect it with each of the individual list of peaks i.e. get the original peaks from each condition that map to every merged peak. That will allow you to map the original summits back to the merged peaks.

--
You received this message because you are subscribed to the Google Groups "idr-discuss" group.
To unsubscribe from this group and stop receiving emails from it, send an email to idr-discuss...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Lucy

unread,
Sep 13, 2018, 8:57:33 AM9/13/18
to idr-discuss
Hi,

Thanks for the response.  I tried the following:
bedtools intersect -a N_optimal.bed -b optimal_merged.bed > test.bed

N_optimal.bed = optimal peak list for one cell type
optimal_merged.bed = global peak list for all three cell types combined

However, when I do this, test.bed comes out exactly the same as N_optimal.bed (which I would expect as all of the peaks in N_optimal.bed will overlap with peaks in optimal_merged.bed).  I am a bit confused as to how I can use this to get the summits for the global peak list.

Best wishes,
Lucy

Anshul Kundaje

unread,
Sep 13, 2018, 9:52:25 AM9/13/18
to idr-d...@googlegroups.com
bedtools intersect -wa -wb -a merged.bed -b sampleApeaks.bed sampleBpeaks.bed 

--
Reply all
Reply to author
Forward
0 new messages