Hi, I have been struggling to get the annotation s...
# raredisease
k
Hi, I have been struggling to get the annotation steps to work (v 2.6.0). The issue is that my annotated VCF files are all empty. I think the issue starts at the step
ANNOTATE_STRUCTURAL_VARIANTS:SVDB_QUERY_DB
because the output file returns no variants, although the input file has plenty of variants that should be annotated. It doesn't throw any errors at this point, but it subsequently causes the pipeline to crash at the genmod rank variants step (guessing because there are no variants to rank?) Below is a table summarising a few variants that are both in the query and reference, and thus should be present in the final file right? Or am I missing something here? Is the issue that the annotation differs between the query/ref? Anyone else seen this problem? This is a test sample
Ah, I think it's related to this git issue https://github.com/nf-core/raredisease/issues/615 because I use gnomad v4.1, and I get that same error message in the log file. But the pipeline keeps going with an empty output vcf in my case
b
you should probably include the error message and also whether you mean just the SVs or also the SNVs
k
Thanks! Yeah so all annotated vcfs are empty, SVs, mobile elements etc. this error in particular is for the SVs, with the message:
Copy code
Error: frequency or hit tag not found! Make sure to set the --in_occ AND --in_frq to the number and frequency of alleles/individuals as presented in the INFO column of the input db
As described in the git issue, I have removed the CNVs from the gnomad file. Hopefully it works now. It was confusing though because the pipeline would disregard this error, and instead crash on
RANK_VARIANTS_SV:GENMOD_SCORE
because the input was empty
b
edit: nvm but for future people reading the temporary fix mentioned is here https://github.com/fa2k/raredisease-configs/blob/main/scripts/gnomad-sv-remove-lines-without-af-an.sh
👍 1