MetaQuast might *silently* fail to detect references in the no-ref mode

[alexeigurevich]

Credits for this bug catch go to Lutz from the Saarland University bioinformatics program, reproduced (with minor edits) below:

It could happen that MetaQUAST does not work as intended and the report.html looks like when just running QUAST. This is a behaviour of MetaQUAST that is suboptimal imho.

When I ran metaquast.py, it ran suspiciously quickly, but from the output, I figured everything went alright since there were no errors nor warnings.

Elapsed time: 0:00:05.208169

NOTICEs: 3; WARNINGs: 0; non-fatal ERRORs: 0

When I opened the report.html it looked like the QUAST output, so I consulted the full log, where I found that there was a problem with the reference genome download. It was due to an error in SILVA which provided the instruction to run the following command:

/home/lutz/miniconda3/envs/mda_module4/lib/python3.12/site-packages/quast_libs/blast/makeblastdb -in /home/lutz/miniconda3/envs/mda_module4/lib/python3.12/site-packages/quast_libs/silva/SILVA_138.1_SSURef_NR99_tax_silva.fasta -dbtype nucl -out /home/lutz/miniconda3/envs/mda_module4/lib/python3.12/site-packages/quast_libs/silva/silva.138.1.db

Which output:

/home/lutz/miniconda3/envs/mda_module4/lib/python3.12/site-packages/quast_libs/blast/makeblastdb: error while loading shared libraries: libidn.so.11: cannot open shared object file: No such file or directory

I figured that there was an issue with libidn11 which I found was not installed on my Manjaro system by default. By installing it, everything worked fine. I think it is a weird behaviour that the output shows no fatal Errors even though MetaQUAST was not able to obtain a reference genome and thus also did not work as intended.