This is a test version of Biostars. For the public version, visit https://www.biostars.org.
bedtools intersect issue

Hi everyone! I have 2 bed files:

A.bed

chr1    1090377 1090381 1       1       +
chr1    1090377 1090381 1       1       +
chr1    1090415 1090419 1       1       +
chr1    1090425 1090429 1       1       +
chr1    1090442 1090446 1       1       -
chr1    1090575 1090579 1       1       -
chr1    1090661 1090665 1       1       +
chr1    1090744 1090748 1       1       +
chr1    1090747 1090751 1       1       -
chr1    1090815 1090819 1       1       +

B.bed

chr1    1090351 1096351 1093351 1       +
chr1    1437425 1443425 1440425 1       +
chr1    1837483 1843483 1840483 1       +
chr1    6792385 6798385 6795385 1       +
chr1    6800919 6806919 6803919 1       +
chr1    8011248 8017248 8014248 1       +
chr1    8061773 8067773 8064773 1       +
chr1    8097782 8103782 8100782 1       +
chr1    8103290 8109290 8106290 1       +
chr1    8104835 8110835 8107835 1       +

and I want to find all intervals that are both in A and B. I used bedtools version v2.26.0 intersect for it:

bedtools intersect -wa -a A.bed -b B.bed

and got:

chr1    1090377 1090381 1       1       +
chr1    1090377 1090381 1       1       +
chr1    1090415 1090419 1       1       +
chr1    1090425 1090429 1       1       +
chr1    1090442 1090446 1       1       -
chr1    1090575 1090579 1       1       -
chr1    1090661 1090665 1       1       +
chr1    1090744 1090748 1       1       +
chr1    1090747 1090751 1       1       -
chr1    1090815 1090819 1       1       +

However, when I tried to force strandedness by adding -s to the command:

bedtools intersect -wa -a A.bed -b B.bed -s

I got no result.

What could be the reason for it?

bed bedtools
• 3,372 views • link
updated by Ram Ram • written by elisheva elisheva

-s parameter in the second command restricts the overlap to the same strand as described in the bedtools intersect manual pages. Are you looking for strand-specific overlaps between two .bed file?

-s  Force “strandedness”. That is, only report hits in B that overlap A on the same strand. By default, overlaps are reported without respect to strand.
• 0 views • link
written by Sej Modha Sej Modha

I know, that exactly what I want and it doesn't seem to work.

• 0 views • link
written by elisheva elisheva

Are you sure that you show us the correct content of A.bed? The coordinates are much smaller than those by B.bed.

fin swimmer

• 1 views • link
written by finswimmer finswimmer

Yes, I know, but that what I want to get eventually.

• 1 views • link
written by elisheva elisheva

??

The both files you show have no overlap. Even without having a look on the strands.

The output of $ bedtools intersect -wa -a A.bed -b B.bed is also empty.

• 1 views • link
written by finswimmer finswimmer

I edited the post, but those are only the first 10 lines, actually, they have overlap, as you can see in the first result.

• 1 views • link
written by elisheva elisheva

Ok, but please show us lines that should overlap. Without it is impossible to test.

BTW: Standard question: What version of bedtools are you using?

• 1 views • link
written by finswimmer finswimmer

I am using version: v2.26.0

• 1 views • link
written by elisheva elisheva

Current version is v2.27.1. But I've tested your updated example data in both version and it works for me

$ bedtools intersect -wa -a A.bed -b B.bed   
chr1    1090377 1090381 1   1   +
chr1    1090377 1090381 1   1   +
chr1    1090415 1090419 1   1   +
chr1    1090425 1090429 1   1   +
chr1    1090442 1090446 1   1   -
chr1    1090575 1090579 1   1   -
chr1    1090661 1090665 1   1   +
chr1    1090744 1090748 1   1   +
chr1    1090747 1090751 1   1   -
chr1    1090815 1090819 1   1   +

and

$ bedtools intersect -wa -a A.bed -b B.bed -s
chr1    1090377 1090381 1   1   +
chr1    1090377 1090381 1   1   +
chr1    1090415 1090419 1   1   +
chr1    1090425 1090429 1   1   +
chr1    1090661 1090665 1   1   +
chr1    1090744 1090748 1   1   +
chr1    1090815 1090819 1   1   +

Do you get an output with the subset example you've posted? If so than your original bed file is corrupted somehow.

• 1 views • link
written by finswimmer finswimmer

That's really weird, and yes, for the example I posted I get an output as I posted either.

• 1 views • link
written by elisheva elisheva

can confirm finswimmer ' s observation with OP data:

   $ bedtools intersect -wa -a a.bed -b b.bed -s
    chr1    1090377 1090381 1   1   +
    chr1    1090377 1090381 1   1   +
    chr1    1090415 1090419 1   1   +
    chr1    1090425 1090429 1   1   +
    chr1    1090661 1090665 1   1   +
    chr1    1090744 1090748 1   1   +
    chr1    1090815 1090819 1   1   +

    $ bedtools  --version
    bedtools v2.27.1
• 1 views • link
written by cpad0112 cpad0112

Even with bedtools version v2.25.0:

$ bedtools intersect -wa -a A.bed -b B.bed -s
chr1    1090377 1090381 1   1   +
chr1    1090377 1090381 1   1   +
chr1    1090415 1090419 1   1   +
chr1    1090425 1090429 1   1   +
chr1    1090661 1090665 1   1   +
chr1    1090744 1090748 1   1   +
chr1    1090815 1090819 1   1   +


$ bedtools  --version
bedtools v2.25.0
• 1 views • link
written by michael.ante michael.ante

0 answers

No answers yet.

Log in to answer this question.

More posts like this