OMArk produces null OMAmer output and IndexError

[Sahar-Abdollahi]

Hi OMArk team,

I’m following the official documentation and testing OMArk with the example dataset.
The OMAmer output file (.omamer) contains only null values. and when I run OMArk, it fails with the following traceback:

WARNING: Database version mismatch: DB 2.0.3 / OMAmer 2.1.1
Traceback (most recent call last):
  File "…/bin/omark", line 52, in <module>
    omark.launcher(arg)
  File "…/omark/omark.py", line 269, in launcher
    get_omamer_qscore(omamerfile, dbpath, outdir, taxid, original_FASTA_file = original_fasta, isoform_file=isoform_file, taxonomic_rank=taxonomic_rank)
  File "…/omark/omark.py", line 43, in get_omamer_qscore
    if "hoglevel" in omamdata[0]:
IndexError: list index out of range

Commands used

omamer search --db  LUCA.h5 --query example_data/UP000005640_9606.fasta --out example_data/UP000005640_9606.omamer

mkdir example_data/omark_output
omark -f example_data/UP000005640_9606.omamer -d LUCA.h5 -o example_data/omark_output

Details

• OMArk and LUCA.h5 were both downloaded according to the documentation.

• The .omamer file is generated but contains only null entries.

• The same behavior occurs with my plant proteomes.

• I’ve tried running it both in a conda environment and with a local installation, with the same result.

Could you please advise what might cause this issue?

[YanNevers]

Dear @Sahar-Abdollahi ,

Thank you for reporting the issue.
I believe this could have been caused by an issue with the OMAmer version that conda installed for you (2.1.1).
This release had a known bug causing all results to take a null value. We released a hotfix for this bug last wekk (see https://github.com/DessimozLab/omamer/releases/tag/v2.1.2).

Could you try updating your OMAmer version in conda to 2.1.2 and see if this problem persists?

Best,
Yannis