Bedtools complement issue

Question

[bioinf_anais]

Hello,

2.25.0 version seems to give wierd result. I get lots of regions with the start higher than the end position. Most of the time here is what I get :
scaffold_ID 1 -1
Is there an explanation to this behavior? I get expected results with 2.19 version so it seems related to the version.
Thanks for any help.

Anaïs

[Istvan Albert]

when you report unexpected results you should show a simple example (one line) for the input and the output you believe to be incorrect as well as the corresponding command line

[bioinf_anais]

You are right. Here is an example.

initial bed file :

scaffold_1055   0       79 (with scaffold_1055 of size 88587)

result of bedtools complement v2.25:

scaffold_1055   0       0
scaffold_1055   80      88586

result of bedtools complement v2.19:

scaffold_1055   79      88587 (expected)

Looks like it is linked to 0-based/1-based issue, doesn't it?

Here is my command line:

bedtools complement -i mybedfile.bed -g mygenome.file

Answer 1

[Alex Reynolds]

Via BEDOPS bedops --complement:

$ sort-bed A.unsorted > A
$ sort-bed B.unsorted > B
$ sort-bed C.unsorted > C
...
$ bedops --complement A B C ... > answer.bed

The output should not contain any data other than the genomic regions between two or more sorted input sets.