Closed mrvollger closed 7 months ago
Adding @holtjma in case he doesn't see these by default.
And thanks to all in advance!
including @sjneph and @adrisede
This sounds like an edge case that wasn't encountered in our initial testing. I'll take a look and verify I can reproduce the issue locally.
Thanks so much! Should only take a couple of seconds, I tried to make the files as small as possible while keeping the error (and let me know if it's not working for you).
I managed to reproduce it locally, will hopefully have a fix in the next few business days.
@mrvollger Turns out the patch was relatively simple, so give v1.2.1 a try on your full file. I verified that the snippet you shared will run to completion now.
Awesome, thanks so much. I will share if I have any issues. Thanks again.
Operating system redhat
Package name hiphase=v1.1.0, pbsv=2.9.0
Describe the bug This isn't a bug per-say, but I do think it will be a common pitfall in default behavior, which may lead to errors.
PBSV converts ambiguous characters to Ns by default re: https://github.com/PacificBiosciences/pbsv?tab=readme-ov-file#why-does-the-vcf-contain-no-ambiguous-iupac-ref-codes However, hiphase checks for exact sequence matches to the reference, so this occasionally results in errors.
Error message e.g.
This causes a W!=N mismatch and error in hiphase.
To Reproduce I have uploaded some small files and the reference is: https://ftp.ncbi.nlm.nih.gov/genomes/all/GCA/000/001/405/GCA_000001405.15_GRCh38/seqs_for_alignment_pipelines.ucsc_ids/GCA_000001405.15_GRCh38_no_alt_analysis_set.fna.gz
Expected behavior Ideally, I think the default PBSV output should work with hiphase. And if not, I would be genuinely curious to hear the reasoning for this design.
Archive.zip