oschwengers / bakta

Rapid & standardized annotation of bacterial genomes, MAGs & plasmids
GNU General Public License v3.0
438 stars 53 forks source link

Bakta Diamond error #272

Closed dannyionescu closed 7 months ago

dannyionescu commented 8 months ago

Hi, I installed Bakta recently using mamba. I downloaded the DB using the built-in command.

I am trying to run mamba on a series of fasta files:

for f in *fasta do sample=${f//.fasta/}

bakta\ --db /project02/DATABASES/bakta/db\ --prefix $sample\ --output $sample.bakta\ --threads 64\ --tmp-dir ./tmp\ $f

done

The program seems to start working correctly producing the following output: parse genome sequences... imported: 1 filtered & revised: 1 contigs: 1

start annotation... predict tRNAs... found: 71 predict tmRNAs... found: 1 predict rRNAs... found: 24 predict ncRNAs... found: 27 predict ncRNA regions... found: 55 predict CRISPR arrays... found: 2 predict & annotate CDSs... predicted: 4061 discarded spurious: 0 revised translational exceptions: 0 detected IPSs: 13

But then it fails producing the following error:

Traceback (most recent call last): File "/project02/miniforge3/envs/bakta/bin/bakta", line 10, in sys.exit(main()) File "/project02/miniforge3/envs/bakta/lib/python3.10/site-packages/bakta/main.py", line 259, in main cdss_psc, cdss_pscc, cdss_not_found = psc.search(cdss_not_found) File "/project02/miniforge3/envs/bakta/lib/python3.10/site-packages/bakta/psc.py", line 64, in search raise Exception(f'diamond error! error code: {proc.returncode}') Exception: diamond error! error code: -11

Bakta is installed for all users (by me) as root in a mamba environment. I tried to run the same command as root but I get the same error.

I ran Diamond test from the bakta environment and all tests (20/20) came back OK.

Looking forward to any tips on how to solve this.

Thanks

Danny

oschwengers commented 8 months ago

Hi @dannyionescu, this is a known bug caused by the recent Diamond version v2.1.9. This has also been reported by other users (https://github.com/oschwengers/bakta/issues/271) and I just manually tested and confirmed this.

This bug is filed upstream https://github.com/bbuchfink/diamond/issues/785. You can downgrade Diamond to v2.1.8 which should solve the issue until this bug is fixed upstream.

valery-shap commented 8 months ago

Hi, I also recently installed Bakta using conda, checked the version of Diamond: it is v.2.1.8. But I still have this issue: `Bakta v1.9.2 Options and arguments: input: AP006726.fasta db: db, version 5.1, full output: test_apo006726 force: True tmp directory: /tmp/tmpiqh5yhpl prefix: apo006726_test threads: 4 translation table: 11

parse genome sequences... imported: 1 filtered & revised: 1 plasmids: 1

start annotation... predict tRNAs... found: 0 predict tmRNAs... found: 0 predict rRNAs... found: 0 predict ncRNAs... found: 1 predict ncRNA regions... found: 1 predict CRISPR arrays... found: 0 predict & annotate CDSs... predicted: 251 discarded spurious: 0 revised translational exceptions: 0 detected IPSs: 248 found PSCs: 2 found PSCCs: 0 lookup annotations... conduct expert systems... amrfinder: 29 protein sequences: 33 combine annotations and mark hypotheticals... detect pseudogenes... Traceback (most recent call last): File "/miniconda3/envs/bakta/bin/bakta", line 10, in sys.exit(main()) File "/miniconda3/envs/bakta/lib/python3.10/site-packages/bakta/main.py", line 302, in main pseudo_candidates = feat_cds.predict_pseudo_candidates(hypotheticals) File "/miniconda3/envs/bakta/lib/python3.10/site-packages/bakta/features/cds.py", line 518, in predict_pseudo_candidates raise Exception(f'diamond error! error code: {proc.returncode}') Exception: diamond error! error code: 1 ` The first launch with another sample run without errors.

Best regards, Valery

Upd. I re-run it with more CPUs and memory and it worked w/o errors.

oschwengers commented 7 months ago

FYI: I temporarily pinned Diamond to v2.1.8 in the Bakta Bioconda package: https://github.com/bioconda/bioconda-recipes/pull/46111

oschwengers commented 7 months ago

This bug is occurring in Diamond and it has been reported upstream. A downgrade to v2.1.8 is currently the best we can do until this is fixed.

As this isn't caused by Bakta but Diamond and can be fixed by a downgrade to v2.1.8, I'll close this for now. If this bug remains, please do not hesitate to re-open this or a new issue.