Question: Can't obtain overlapping Peaks
0
gravatar for VY
21 months ago by
VY120
London
VY120 wrote:

Hello,

I've generated peaks using MACS2 for H3K4me3 and H3K27me3 and now want to find the overlapping regions. I'm doing this as I use to (back when using MACS1) by going to: Operate on Genomic intervals, and then using the intersect function to retrieve the overlapping intervals.

Both of my data sets appear in the drop down menu and I set the overlap to be of 1bp minimum but once the job runs the resulting file is empty. I've looked at the bed peak files in the genome browser and can see the peaks overlapping at certain regions so not quite sure why my job returns back to me empty.

Any suggestions? Cheers

gops interval macs2 chip-seq • 530 views
ADD COMMENTlink written 21 months ago by VY120

Are you working at http://usegalaxy.org? If so or can reproduce the problem there, we can take a look at the inputs and job parameters to see what is going on. My concern is that there may be a tool issue - and if there is - we want to capture it and get it fixed.

First, be sure the inputs and this result job are undeleted in the working history. Next generate a history share link (how-to). Then email that share link to galaxy-bugs@lists.galaxyproject.org. In the email also note the problem output dataset number and please include a link to this Biostars post.

Thanks, Jen, Galaxy team

ADD REPLYlink written 21 months ago by Jennifer Hillman Jackson25k
1
gravatar for VY
21 months ago by
VY120
London
VY120 wrote:

Hi Jen,

Did some digging in my datasets and the issue doesn't seem to be with the intersect tool afterall but with MACS2. For some reason if I look at the attributes of my "summits in bed" MACS2 output. The column labels (start, end, strand etc) and the column numbers automatically detected don't seem to match for some reason. If I manually go through my peak files and use autodetect in each one of them it corrects this and then the intersect function works.

Not sure why MACS2 would these wrong when generating the files.

best wishes, Val

ADD COMMENTlink written 21 months ago by VY120

Odd result - I'll see if I can reproduce. It is possibly related to a known metadata issue with BED files (ticket here: https://github.com/galaxyproject/galaxy/issues/2175).

Thanks for posting back the solution you found! I'll promote that to the official "answer".

Jen, Galaxy team

ADD REPLYlink written 21 months ago by Jennifer Hillman Jackson25k
Please log in to add an answer.

Help
Access

Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
Powered by Biostar version 16.09
Traffic: 84 users visited in the last hour