search

Dfam-consortium / RepeatModeler

De-Novo Repeat Discovery Tool
Other
182 stars 23 forks source link

RepeatModeler ended with: famdef failed. Exit code 768 #214

Closed kdarragh1994 closed 10 months ago

kdarragh1994 commented 10 months ago

Hi,

I am trying to run RepeatModeler on a plant genome for which no repeat library exists. It ended after two days with the following errors:

The slurm file shows the error: "famdef failed. Exit code 768".

This file also contains thousands of error messages in all rounds: _"In file: /RM21209.TueAug221747112023/round-4/summary/eles. Ignoring line and moving on. WARNING: RECON returned a negative offset:"

I looked at the logs of the run and it ended in round 5 but no further error message is detected. In round-5/tmp/log2 the final line is: _"evaluating family membership of element 4508 error: ele 4508 reversed from readin"

Command given to program:

module load repeatmodeler/2.0.3 BuildDatabase -name lasius.DB -engine rmblast GCA_027563935.1_ASM2756393v1_genomic.fna RepeatModeler -database lasius.DB -engine ncbi -pa 16

This was run on the Berkeley HPC cluster so the module was not installed by me.

Thanks so much for any advice on what to try!

rmhubley commented 10 months ago

I am afraid this is a problem with the RECON program that RepeatModeler invokes. This type of bug is rare and hard to debug. I am happy to look into it but if you want to move forward with your analysis I would recommend re-running your analysis from the start. RepeatModeler uses a random sampling technique and it's extremely unlikely it would generate the exact set of sequences that caused the problem for RECON in a second run. To look into the exact problem, we would need a copy of the output directory (RM_21209.TueAug221747112023 in your case).

kdarragh1994 commented 10 months ago

Thank you for your response. I tried running with singularity and it worked first time so I think was an issue with the installation on the HPC!