Option to Skip the AMR detection step

[abelTan131]

Is your feature request related to an existing issue or bug?
Briefly reference any existing issues this is related to.

I run my file and got:
Traceback (most recent call last):
File "/opt/conda/bin/bakta", line 8, in
sys.exit(main())
File "/opt/conda/lib/python3.8/site-packages/bakta/main.py", line 275, in main
expert_amr_found = exp_amr.search(cdss, cds_aa_path)
File "/opt/conda/lib/python3.8/site-packages/bakta/expert/amrfinder.py", line 47, in search
raise Exception(f"amrfinder error! error code: {proc.returncode}. Please, try 'amrfinder_update --force_update --database {amrfinderplus_db_path}' to update AMRFinderPlus's internal database.")
Exception: amrfinder error! error code: 1. Please, try 'amrfinder_update --force_update --database /tllhome/abel/bakta/db-light/amrfinderplus-db' to update AMRFinderPlus's internal database.

Is your new feature related to a general problem?
Please provide us with a concise description of what the problem is:

• what happened which was not expected
• what did not happen which was expected
• what would you like to happen or see in the output
  I dont want the AMR genes, I just want the gene annotation. Can we skip it?

Describe the solution you'd like
A clear and concise description of what you want to happen.
Option to skip the AMR detection step, so I dont get this error.

Describe alternatives you've considered
A clear and concise description of any alternative solutions or features you've considered.

Additional context
Add any other context or information about your feature request here.

[gracehm]

Hi, I just wanted to add that we are also facing the same issue, only as of yesterday. We haven't been able to solve it yet, but I wanted to add in case it is a potential update issue.

[oschwengers]

Hi @abelTan131 , @gracehm ,
thanks for reporting. We also see these issues with the AMRFinderPlus update from time to time. However, I'm still puzzled what causes these effects. I have updated the DB download routine and relaxed file permissions which are set during the download/update. Maybe this helps a bit here. This is active in the new v1.9.2 patch release. I also published a huge database update this week v5.1. You might want to try this one.

FYI: the amrfinder_update --force_update --database <path-to-bakta-db>/amrfinderplus-db/ command as suggested by the Bakta error output did the job in my cases.

[aldertzomer]

Same problem here. Will try the update. Does it have to do with this amrfinderplus issue: ncbi/amr#89 ?

[oschwengers]

Hi @aldertzomer, sorry for the late reply. No, the issue you've referenced is quite old and is already fixed IMHO. So, I don't think that this error is related to that issue. I rather see this caused by the new AMRFinderPlus DB version which is not compatible with former versions. Hence, an update of Bakta with the newest AMRFinderPlus version and an DB update should solve this.

[DanielleMStevens]

I was having the same issue. My fix was updating both Bakta (v1.9.3) and AMRFinderPlus with both:

amrfinder_update --force_update --database <path-to-bakta-db>/amrfinderplus-db/
amrfinder --force_update

Though the first few tries didn't work until I updated the read/write/execute permissions of the amrfinderplus-db directory. Can't say if its the same issue but may be something to check.

[sariamead]

I am having the same issue. I've reinstalled Bakta via conda (v1.8.1) and tried the force update for AMRFinderPlus. Bakta seems to work fine until reaching the AMR detection step and then throws the following error:
Exception: amrfinder error! error code: 1. Please, try 'amrfinder_update --force_update --database db_path to update AMRFinderPlus's internal databases

[oschwengers]

Could you please try to update Bakta and AMRFInderPlus to the latest versions and then re-install the Bakta db?

It is certainly related to different Software/Database version dependencies of AMRFinderPlus.

A successful quick test:

$ mamba create -p ./conda-env --quiet --yes bakta=1.9.4
Preparing transaction: ...working... done
Verifying transaction: ...working... done
Executing transaction: ...working... done

$ mamba activate ./conda-env/
(/home/oliver/tmp/bakta-test/conda-env) $ bakta_db download --type light
Bakta software version: 1.9.4
Required database schema version: 5
Selected DB type: light
Fetch DB versions...
	... compatible DB versions: 2
Download database: v5.1, type=light, 2024-01-19, DOI: 10.5281/zenodo.10522951, URL: https://zenodo.org/record/10522951/files/db-light.tar.gz...
|████████████████████████████████████████| 1.48G/1.48G [100%] in 40.7s (36.46M/s) 
	... done
Check MD5 sum...
	...database file OK: 31b3fbdceace50930f8607f8d664d3f4
Extract DB tarball: file=<...>/db-light.tar.gz, output=<...>bakta-test
Successfully downloaded Bakta database!
	version: 5.1
	Type: light
	DOI: 10.5281/zenodo.10522951
	path: <...>db-light
Update AMRFinderPlus database...
	... done
Run Bakta using '--db <...>db-light' or set a BAKTA_DB environment variable: 'export BAKTA_DB=<...>/db-light'

$ bakta --db db-light/ --output test ~/software/bakta/bakta/test/data/NC_002127.1.fna
parse genome sequences...
...
start annotation...
...
predict & annotate CDSs...
	predicted: 3 
	discarded spurious: 0
	revised translational exceptions: 0
	detected IPSs: 0
	found PSCCs: 3
	lookup annotations...
	conduct expert systems...
		amrfinder: 0
		protein sequences: 2
	combine annotations and mark hypotheticals...
	revise special cases...
extract sORF...
	potential: 52
	discarded due to overlaps: 36
	discarded spurious: 0
	detected IPSs: 0
	found PSCCs: 0
	lookup annotations...
	filter and combine annotations...
	filtered sORFs: 0
...
Annotation successfully finished in 0:15 [mm:ss].

[StickHu]

Same problem here. I've using the command line 'amrfinder_update --force_update --database /data/workdir/huwa/annotation/bakta_annotation/db/amrfinderplus-db'

But it failed with that

*** ERROR ***
CURL: Cannot read
from https://ftp.ncbi.nlm.nih.gov/pathogen/Antimicrobial_resistance/AMRFinderPlus/database/
code=6
error: Could not resolve host: ftp.ncbi.nlm.nih.gov
version: 8.8.0

HOSTNAME: localhost
SHELL: /bin/bash
PWD: /data/workdir/huwa/annotation/bakta_annotation/db
PATH: /data/workdir/huwa/software/miniconda3/envs/bakta/bin:/data/workdir/huwa/software/miniconda3/condabin:/home/huwa/.local/bin:/home/huwa/bin:/usr/share/Modules/bin:/usr/local/bin:/usr/bin:/usr/local/sbin:/usr/sbin:/data/workdir/huwa/software/miniconda3/bin:/data/workdir/huwa/software/miniconda3/bin
Progam name: amrfinder_update
Command line: amrfinder_update --force_update --database /data/workdir/huwa/annotation/bakta_annotation/db/amrfinderplus-db`

Could you please give me some advice on it? Thanks

[oschwengers]

Hi @StickHu ,
this seems to be a generic issue on your site. Looks like your machine cannot access the NCBI FTP server at all. Could you ask a local admin to help out?

[zexlibioinfo]

Hi @oschwengers,

I get a same update error on amrfinderplus database. I have tried to either reinstall bakta, or force update database, or directly download latest database from the amrfinderplus GitHub page but unfortunately can't resolve the issue. It is worthy to mention that I get a new error message running with debug mode: Protein sequences looks like a nucleotide sequences.

Here is the command I used:

bakta --db <path/to/bakta_db> --output ./bakta_test --prefix test --threads 32 --debug ./nucleotide/NZ_AAKL01000181.1.fna

Attach the test sequence:

>NZ_AAKL01000181.1
ACCTCCTCNNAAACATTAACTAACATCTCATATTTANTGNAANTTTNNGGNTTTTTTCCC
GATTTATGNTAGCCAAGAATCCTTTATATCCTTTCAGCATTAATATCAATATTCCAAAAT
GGTCCTCTTCATTCTCCAAGCCCCGNANNTCNNNTTCNNCNNNCNNNCCCCCCTTTTTTG
TCTCTTCNNCNTCCCNNTCTTTTTTCTCCCCTTNTACTAGACAGNGTNNNNNCNNNNNNN
CCTCCCTTCTTTCGTCTTTCCTCTTTTCTCCTTCCTNTCCCCCTTTTCTGGGGGGGGAAG
CGCGGCTCCTTTTTTCCCCCCCTCTTCCCCCTNCCCGTTCTCTTCCTGTGCTCCTTCCTC
CCTCCCTTTTCCCCCTGCCCTCGCCTTTCCGGCCTCCTCCCCTCCCCGTGTCCGCTCCCT
TTCCTCCCGCAACCGGCTTTCTCGCTCCCCTCTCCTCCCTCTCCCCTTCCCCCCCTCCTC
TTTGCTTGTCTCTCTCCCTCCTCCTCCCCTCTCTCCCCCCCCTCTTTCCTCTTTCTCCTT
TTTTCTCTCCTCCCTTCTCCCCCTCCCCCCTTTTGTGGTGGGGGGGGGGGGAAACCCCCC
CTTTTTTTTTTTTTTCCCCCCCCCCCCCCCNNNNNNNNNNNNNNNCNNCTTTTTAAAAAA
AAAAAAAAGAGGGGGGGTCTTCTCTTATCCTCCTCCACACACTATACTAAAAACATCAAC
AATATACAATAAATAAATCCATCCAAACAACATCACTACTTATAATAAACTACTAATTCT
AACAACCCCCACAACAACCACACACACACCCTACTAAACTCATAACTACCACACAATACT
CATACACTCTCACCATAAGAACAATAACTCAAATACTACATAAAAAAACAAACATATCCA
TCACTACAACTAATAACACCACACTAAAATATACATAACACCAATCAAAAATACCCACCC
TAATTCTAATACCCCACAAATAACTACTATTTACCAAACCCAACAACCAAAAATACCCAT
ATATTCTAAAAACTTCATACCCAACTAAACTAATCACAAACAAACAATCACACACAACCA
ACACTATCCAACCATCATTTTCCCACCCATATTAACTACTCCATCAAACATAAATACACT
AACACACAATCTCCCATCAACACCACATCCTTCACACAAACAAATCACAAAACTTAACCA
ACTCAAATATCAATTACAACTTCTACCATTATATTCATACACCAAATAACATCTCAACAT
CCATATCTCTCCTATCTAACTATCACATAAAAATAATTCAAATAAAAATATAAACCAAAT
ACTTCCAATTCTCCTCCCTACACCCATCCACTACCTATACCATCAATATCACACTCATAT
TAAAACACAAAACTCCATAACACACCACTCATATAATATTCAAATCTAAAACATTCTTCT
CATAAACAACCAACACAATCCACTAATAATAAACAATCTACTCACAAATTCTACCATCAT
ACCAATACATTACCCTTATACTACCCACTATTTCAAAACCTCTATCCCTAACTACCATAC
TTACTCATATCTCCATACCATCCCTTCACAATTNACATACCAACTATACACACATCTATC
ACATTATTACAATACCACAAATCAACACAATCACACCAATTGTCTCATAACTACACCACC
TTTAACATATTCTACAGATCATTCACATA

[ppflrs]

Thanks @oschwengers for your work in bakta! Is there any chance to get the option to skip AMR detection added in future releases?