#27 Alignments sometimes off by one?

[Dan Bolser]

We've found that some ATAC alignment start at 0, while some start at 1. In all the cases we looked at where the alignment started at 1, the first base was a match (i.e. it /should have/ started at 0).

We don't know if this happens elsewhere, i.e. an alignment starting at 1023345 when it should start at 1023344.

Cheers,
Dan.

[Brian Walenz]

I will guess this is a reverse complement base-based vs space-based problem.

I won't go hunting for the problem, but will commit patches and/or give advice.

[Dan Bolser]

Thanks Brian, by space-based, do you mean zero offset?

We aren't seeing this uniquely on reverse complement alignments. Here is an example:

M u H100255 r4139 IWGSC_CSS_3B_scaff_7104925 1 202 1 3B 185959320 202 1
M u H100257 r6175 IWGSC_CSS_3B_scaff_8869715 1 202 1 3B 106962489 202 1
M u H100259 r13430 IWGSC_CSS_3B_scaff_9132968 1 203 1 3B 122008361 203 1
M u H100262 r23056 IWGSC_CSS_3B_scaff_7815390 0 189 1 3B 736502984 189 -1
M u H100263 r44865 IWGSC_CSS_3B_scaff_7974030 1 209 1 3B 115961908 209 1
M u H1841859 r45350 IWGSC_CSS_3B_scaff_9796224 0 211 1 7B 210653543 211 1
M u H100265 r45655 IWGSC_CSS_3B_scaff_7104447 1 211 1 3B 567415299 211 -1
M u H100268 r45762 IWGSC_CSS_3B_scaff_7657584 1 211 1 3B 150847520 211 -1
M u H100270 r45966 IWGSC_CSS_3B_scaff_873520 0 213 1 3B 538477585 213 -1

In all cases where the start is 1, it should be 0.