How does "bedtools intersect -s" work with paired-end sequences?

Question

[glyptonotus]

Hi,

I have paired-end reads from RNAseq experiments and I want to use bedtools intersect with the force “strandedness” parameter (-s). Will bedtools take into account the paired-end reads and treat them as a single event, or will get two lines of output, one for each of the two pairs?

Thank you!

[joseemayo2]

I ran into a similar issue today. I do not have a direct answer to your question, but a work around that I found is to combine the mates into a bedpe file, which will keep the coordinates of each mate and you can parse them as needed. The advantage is it's a simple way to force the pairs as a single entry. This post is what sparked the idea. Hope this helps!

Answer 1

[glyptonotus]

In case anyone is interested, I got the answer: bedtools intersect -s discriminates between strands in paired-end reads. The way I checked it:

I generated a simple bed file with just two rows with identical start and end coordinates but different strand (BED file):

chr1    3214481 3671498 AAA4    1   -

chr1    3214481 3671498 AAA4    1   +

After doing:

bedtools intersect -a (BED) -b (BAM: paired-end alignments) -s > output.bed

output.bed:

chr1    3214481 3671498 AAA4    1   -   427

chr1    3214481 3671498 AAA4    1   +   439

I obtained the same number of counts in the two strands, thus bedtools is not considering paired-end reads as a single event, but it assigns each read of the pair to one different strand.

How would you solve this problem?