luntergroup / octopus

Bayesian haplotype-based mutation calling
MIT License
299 stars 37 forks source link

Segmentation fault during trio variant calling #199

Open PfeiferLab opened 2 years ago

PfeiferLab commented 2 years ago

Hi,

Similar to one of the previous tickets (Segfault occurring during somatic variant calling #50), I am experiencing a Segmentation Fault for certain chromosomes when trying to run Octopus in the Trio mode, e.g.

[2021-07-07 01:40:00]  <INFO>  chr1:113388703  51.4%  12h 17m  11h 37m
Segmentation fault

Version:

octopus version 0.7.4
Target: x86_64 Linux 5.4.0-72-generic
SIMD extension: AVX2
Compiler: GNU 9.3.0
Boost: 1_74
(installed via Anaconda)

Command line:

octopus \ 
-R reference.fa \ 
-I sire.bam damn.bam offspring.bam \ 
-M damn \ 
-F sire \ 
--snp-heterozygosity 0.002 \ 
--denovo-snv-prior 1e-8 \ 
--forest germline.v0.7.4.forest \ 
--sequence-error-model PCR-FREE.HISEQ-2500 \ 
-o trio_chr1.vcf.gz \ 
--threads 28 \ 
-T chr1

Changing the memory size results in different errors (though the debug log doesn’t contain any errors). Happy to share the .bam and .fasta files via GoogleDrive or Dropbox in case it might be helpful!

Any help would be greatly appreciated!

dancooke commented 2 years ago

Thanks for the bug report. Probably the quickest way to get this resolved is if you provide me with the BAM and FASTA files. Either Google Drive or DropBox works.

PfeiferLab commented 2 years ago

Thank you for looking into this, @dancooke ! I emailed you a link to the .bam and .fasta files.

jbedo commented 2 years ago

I ran into a segfault recently #228, might not be realted as this is with the trio model. I've patched the segfault in #229 to throw error messages instead of segfaulting.