luntergroup / octopus

Bayesian haplotype-based mutation calling
MIT License
305 stars 38 forks source link

Job killed due to `ContigRegion: expanding past contig start` #244

Open ja-Reeve opened 2 years ago

ja-Reeve commented 2 years ago

I'm testing out Octopus for a WGS dataset for 300 sea snails (Littorina saxatilis). I trialed Octopus on a single individual, but after running for 3 days the job stopped when it encountered a ContigRegion: expanded past contig start error.

How should I fix this bug?

Version: 0.7.4

Script:

$ octopus \
        -R $DIR/ref_genome/contig_level_assembly.fasta \
        -I $DIR/bam/Test_snailWGS.bam \
        -o $DIR/vcf/Test_snailWGS.vcf.gz \
        --threads 10

Error:

[2022-09-20 10:25:08] <EROR> Encountered a problem whilst calling Contig116055:0-4120(ContigRegion: expanded past contig start)
[2022-09-20 19:35:00] <INFO> Removed 777240 temporary files
[2022-09-20 19:35:03] <INFO>               -             100%           3d 13h                 -
[2022-09-20 19:35:06] <EROR> A program error has occurred:
[2022-09-20 19:35:06] <EROR>
[2022-09-20 19:35:06] <EROR>     Encountered an exception during calling 'ContigRegion: expanded past
[2022-09-20 19:35:06] <EROR>     contig start'. This means there is a bug and your results are
[2022-09-20 19:35:06] <EROR>     untrustworthy.
[2022-09-20 19:35:06] <EROR>
[2022-09-20 19:35:06] <EROR> To help resolve this error run in debug mode and send the log file to
[2022-09-20 19:35:06] <EROR> https://github.com/luntergroup/octopus/issues.
[2022-09-20 19:35:06] <INFO> ------------------------------------------------------------------------

Update: It looks like this is a specific issue with the sample I used. I tried calling the problem region on a different bam, and it worked. Update 2: Further testing found this same error in 3/5 samples.

Cheers, James