Entering edit mode
7.0 years ago
kerrypop
•
0
Hello,
I am trying to use bedtools to find intersections between H3K4me1 peaks and my regions of interest. Though I can see overlap on IGV, intersect returns 0 without any error messages.
Here is command I'm using and screenshot from IGV:
bedtools intersect -a gene.ROI -b wgEncodeBroadHistoneHuvecH3k4me1StdPk.broadPeak -u | wc -l
IGV screenshot: https://imgur.com/a/JYUzp
Is there something I'm missing?
Thank you, Kerry
do file use the same chromosome notation ? are the files sorted ? are both files BED files ?
Yes, same chromosome notation. Files are not sorted - I've done this analysis before and didn't sort them. Do you think sorting would overcome this? The ROI file is just a document, created using nano in terminal. The other is a broad peak file.
Check columns are tab separated and there are no funny, non-printable characters. If you do something like:
You should see columns separated by
^I
, e.g.:I only used chromosome coordinates but don't see any extraneous characters
It looks like the dollar sign at the end of each line might be causing trouble. You can use this command to remove the last character from each line:
That is because of options (
-vet
) used forcat
command.Your IGV screenshot seems to show the entire genome. It may be that once you zoom in into one of the ROI genes there is actually no intersection.
Here's zoomed in screenshot of the first ROI. There should be 4 intersections. https://imgur.com/a/lsiEQ