Hi MAGgers. I have a question regarding prokka which is taking quite a bit of time to finish. I run the MAG workflow using the coassembly option (all samples as one group) but I realised that DAS refinding output is not carried through to the annotation stage by default so I also flagged. see command below. Prokka is taking long. Is there a way one can speed up the annotation.

nextflow run nf-core/mag -r 4.0.0  -profile kaust -resume -c mag.cfg \
	--input samplesheet.csv \
	--outdir results \
	--gtdb_db ./gtdb-tk/release220/ \
	--skip_spades true --skip_concoct true \
	--prokka_compliance_centre NCBI --prokka_with_compliance true \
	--refine_bins_dastool true \
	--binqc_tool checkm2 \
	--coassemble_group true –binning_map_mode all \
	--postbinning_input both

Prokka just takes a long time, there isn't really a way around it, unless you simply skip it. I'm not sure what you mean by DAS_Tool output doesn't go all the way through though?

--postbinning_input both

i meant that the default behaviour doesn’t carry forward the polished bins.

Ah, so you mean you cancelled and resumed setting

--postbinning_input both

?

No, I didn’t cancel. The initial job finished default

--postbinning_input

then i resubmitted with

--postbinning_input both

but prokka is taking too long.

Ah ok

Bakta? perhaps I can use bakta myself then 😉

If you're only interested in Bacteria: https://github.com/oschwengers/bakta

nice, thanks for the tip