Assemble step error

[poddarharsh15]

Hello, I am trying to run GRIDSS on BAM files, however during the process it gives me some errors that I couldn't figure out could you please suggest me some advice? Thanks.

gridss.full.20231115_110358.flgambardella01.256857.log

- [ ] [ ] gridss -r /home/tigem/h.poddar/input/reference_genome/GRCh38.primary_assembly.genome.fa -j /home/tigem/h.poddar/anaconda3/envs/GRIDSS/share/gridss-2.13.2-2/gridss.jar -t 8 -s assemble -a ~/input/GRIDSS/assembly.bam -w ~/input/GRIDSS  --jvmheap 15g --otherjvmheap 15g --skipsoftcliprealignment /home/tigem/h.poddar/input/lumpy/sample_sorted.bam
- [ ] [ ] Using working directory "/home/tigem/h.poddar/input/GRIDSS

[d-cameron]

it gives me some errors that I couldn't figure out could you please suggest me some advice?

There's no ERROR messages in that log. That output is what you'd expect. What isn't expected is the GRIDSS run starting with

INFO 2023-11-15 11:03:59 AssemblyEvidenceSource Starting assembly on chunk 14

That log looks very much like you'd previously run GRIDSS, the job was killed (probably ran out of your time allocation on your HPC cluster), and GRIDSS was restarting where it left off.

Give your GRIDSS job more time before killing it. It's running, but the job is getting killed before it can finish.

[d-cameron]

It could also have been killed by running out of memory you set --jvmheap 15g --otherjvmheap 15g but the default recommended memory is 31Gb. If you don't have 32Gb of memory on the machine you're running it on then change -t 8 to -t 4 so you're only running 4 threads at once instead of 8. This will approximately halve your memory usage so it should fit within the 15gb you've given it.

[poddarharsh15]

Thank you for answering, I tried to run the pipeline again with 4 threads but still it's get killed after couple of minutes. gridss.full.20231120_132758.flgambardella01.26468.log Could you please check these results?

[d-cameron]

Exit code 137 indicated killed due to out of memory. You'll need to reduce thread count further or use a machine with more memory.

On Mon, Nov 20, 2023 at 11:37 PM poddarharsh15 @.***> wrote:

Thank you for answering, I tried to rub the pipeline again with 4 threads but still it's get killed after couple of minutes. [image: Screenshot from 2023-11-20 13-34-18] https://user-images.githubusercontent.com/45700858/284267379-3ed52f8c-a4e6-49a1-aabb-91c39fd632a9.png gridss.full.20231120_132758.flgambardella01.26468.log https://github.com/PapenfussLab/gridss/files/13414636/gridss.full.20231120_132758.flgambardella01.26468.log Could you please check these results?

— Reply to this email directly, view it on GitHub https://github.com/PapenfussLab/gridss/issues/648#issuecomment-1818987223, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABOBYOAIJ63WT775U2OVKJTYFNFIHAVCNFSM6AAAAAA7MHUWDWVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMYTQMJYHE4DOMRSGM . You are receiving this because you modified the open/close state.Message ID: @.***>

[poddarharsh15]

Hello again, I reduced the thread count to 2 and then 1 but still gives the same error with 137 exit code I assume I need to run the GRIDSS on cluster to provide more memory and time to run.