-
Notifications
You must be signed in to change notification settings - Fork 37
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Starting Call Set Refinement (CSR) filtering provokes RAM overload #280
Comments
Hi, I'm currently having similar issues with Octopus processing human whole genomes (30x) using the docker image and singularity. Octopus runs well until the Call Set Refinement filtering is started, after which it is killed by an oom leak. I haven't been able to figure out yet how to solve this. I have tried multiple configurations, adjusting the maximum RAM memory (up to 250Gb, 12Gb per core) without any success. Any ideas? Version Command Thanks in advance for your time! |
In the past, I have only run octopus on 30x WGS human with something like 1900G RAM, but I did not look at the peak RAM usage. I would recommend you split your job into parts, and maybe try again per chromosome. Do you really have the forest model file? That is something that could also cause an error if you do not have it, but I am not sure if it would cause the same OOM error. |
Hi.
I'm trying to analyze data from viruses (in illumina paired end fastq format), with a size of 100 MB for the sum of R1 and R2, in a machine with 16 cores and 30 GB of ram memory.
In previous steps, fastp tool is used to filtering and trimming adapters from raw data, and bwa mem to generate aligned cram file.
When Starting Call Set Refinement (CSR) filtering section is reached, RAM is growing up to the 100% of load capacity and kill the process abruptly.
https://filesender.uvigo.es/?s=download&token=4bced40f-f8cd-4930-a373-fc68e59e7a64
Version
octopus version 0.7.4
Command
vhd.run.cloud.sh: line 82: 72876 Killed
$ octopus --threads $THREADS -P 1 -R $VIRUSREF -I $SAMPLE.CRAM --annotations AD ADP AF --sequence-error-model PCR-free.HiSeq-2500 -o $SAMPLE.OG.vcf.gz
Additional context
Previously, another sample of same type of virus, but with a size of 33 MB for the sum of R1 and R2 fastq files, was completely done, with no errors.
Thank you very much for your support.
The text was updated successfully, but these errors were encountered: