bed format, START always bigger than END, no matter which strand?

Question

[bxia]

I am trying to convert bowtie to bed format.

in bowtie, it only provide start position, if the strand is '-', is that the START in bowtie is actually the left of the reads, no matter which strand?

And in bed format, START always bigger than END, no matter which strand?

For example,

in bowtie I have

read_name - chr1 714602 CCTGA ##### 0

when I covert to bed,

it should be

chr1 714602 714607 .....

OR

chr1 714598 714602 .....

OR

chr1 714602 714598 ....

Thanks

[GenoMax]

Just to be sure, are you referring to converting a BAM file generated by bowtie to BED?

[bxia]

I am trying to convert .bowtie file generated by bowtie to BED

Answer 1

[Alex Reynolds]

According to the bowtie documentation, the offset is 0-based and relative to the forward strand. In a typical usage of BED formatted data, the start position of an interval is always one or more bases less than the stop position. So I'd pick chr1:714602-714607, based on my interpretation of the documentation and the format specification.

It seems like you could do a sanity check by taking the reverse complement of the sequence and seeing if it matches with the (forward strand-oriented) reference genome at 0-based position 714602. If I'm wrong, then adjust coordinates accordingly (and please follow up here -- that part of the documentation seems a little ambiguous, perhaps!).

[bxia]

That is a good idea. Thanks

[GenoMax]

Go ahead and accept the answer (green check mark) if it solved your question.