AnnotateInsertedSequence does not respect --externalaligner commandline argument

[tgong1]

Hi Cameron,

I'm running GRIDSS v2.12.2 on HCC1395/HCC1395BL. BAM files are around 100x downloaded from:
ftp://ftp.ncbi.nlm.nih.gov/ReferenceSamples/seqc/Somatic_Mutation_WG/data/SPP/SPP_GT_0-1_1.bwa.dedup.bam
ftp://ftp.ncbi.nlm.nih.gov/ReferenceSamples/seqc/Somatic_Mutation_WG/data/SPP/SPP_GT_1-0_1.bwa.dedup.bam

While I've experienced super long time >48hrs to finish the step of PreprocessForBreakendAssembly, which I've never experienced before.

Here is the log file, hope you can give me some suggestions.
gridss.timing.20211027_215025.comput15.73448.log

gridss.full.20211027_215025.comput15.73448.log

Thank you for the help,
Tingting

[d-cameron]

Looks like some sort of failure of the streaming alignment. Try rerunning with --externalaligner

[tgong1]

Hi Daniel,

Thank you very much for the answer. It was running smoothly, while I now find it stops at the "Running AnnotateInsertedSequence" step.

Please see the attached log files. Could you give me some suggestions to solve this?
gridss.full.20211115_150010.comput24.163594.log

gridss.timing.20211115_150010.comput24.163594.log

Thank you for your time.
Tingting

[d-cameron]

Definitely looks like an issue with the in-process bwa aligner in your environment because that's the final place it's using it.

It seems I overlooked external alignment on that set. I'll work on a fix but in the meantime, you have a usable VCF in gridss_SPP_GT.sv.vcf.gridss.working/gridss_SPP_GT.sv.vcf.allocated.vcf (that's just missing RepeatMasker annotations).