Should Bowtie Hits On Negative Strand Be Counted Downwards To Find The End Of The Alignment (Illustrated)?

Question

[Click downvote]

Here I have shown two hits from bowtie (in red), both at position 1000, with a read of length 7. However, one is on the negative strand and one is on the positive.

The negative hit would be reported like this:

0 - chr14:800-1927 200 AAATGTG

What I am wondering is whether the negative strand hit spans 1000-1007 or 993-1000. The latter option is shown in my illustration.

[Istvan Albert]

you need to specify what format are you considering here - in all cases the format definition is the standard the adhere to - you should never just call it a "bowtie hit" - that is gateway to long term confusion

In general representing the 5' end of the reads is more useful in interpreting the data people often transform from SAM format to a (custom) format that allows an easier extraction of that that information. Still it is important to properly define these and not just use generic terms.

Answer 1

[Ido Tamir]

The holy scripture of the SAM says: 1.4.4:

POS: 1-based leftmost mapping POSition of the first matching base. The first base in a reference sequence has coordinate 1. POS is set as 0 for an unmapped read without coordinate. If POS is 0, no assumptions can be made about RNAME and CIGAR.

And the heretics of the Bowtie manual tend to say to this important question:

0-based offset into the forward reference strand where leftmost character of the alignment occurs

And while they seem to agree, in the general view of giving the position of the left on the forward strand, even if the read is in the reverse position, they fundamentally differ in the representation of the world projected onto a coordinate system.

But whoever tries to interpret his results in the light of proprietary file formats and not in the light of the SAM, should forever rot in hell.

PS: -S is your friend

[Click downvote]

Thanks for your answer. I take it to mean using -S means I can count upwards since the POS explanation does not mention strand.