I would rather these to go in the official module @d4straub but I'm around for fast review for that too

Looking good so far, just I would request porting this to the official module (like I said above, I'm available for rapid review on that)

Thasnks, PR to modules is open.

I was just thinking @uel3 mentioned this issue in the context of co-assembly, do we need to distinguish the mechanism being added here between standalone or co-assembly?

Not sure I understand the question. In my tests this works with and without co-assembly, it depends rather on --binning_map_mode, and it works in my tests as expected.

Looking good so far, just I would request porting this to the official module (like I said above, I'm available for rapid review on that)

Thasnks, PR to modules is open.

OK will look now!

I was just thinking @uel3 mentioned this issue in the context of co-assembly, do we need to distinguish the mechanism being added here between standalone or co-assembly?

Not sure I understand the question. In my tests this works with and without co-assembly, it depends rather on --binning_map_mode, and it works in my tests as expected.

That answers my question basically 👍

@d4straub module can be installed in this PR now :)

Just going to run a test with current version and your changes to compare, but code looks good 👍

@d4straub I had a look and it seems to run through, however when I run with test_virus_identification which does co-assembly, the depths/bin_summary.tsv only reports CONCOCT... do you see that in your test run also?

@jfy133 I checked the code and CONVERT_DEPTHS is only channeled into ch_maxbin2_input which only goes into MAXBIN2. As far as I see, of the many files that are available from MAXBIN2.out, summary file isnt used, and that file is the only file that changes in this PR (MaxBin2's summary file looses column that has abundance with the fix).

The following works as expected for me (this runs on this modified branch), i.e. GenomeBinning/depths/bins/bin_depths_summary.tsv & GenomeBinning/bin_summary.tsv contain all bins:

nextflow run mag -profile test_bbnorm,singularity --skip_binning false --centrifuge_db false --kraken2_db false --outdir results_test_bbnorm_3-1-0dev -resume --skip_concoct
nextflow run mag -profile test_bbnorm,singularity --skip_binning false --centrifuge_db false --kraken2_db false --outdir results_test_bbnorm_3-1-0dev_concoct -resume

test_virus_identification doesnt produce any binning results, because it uses

    skip_binning                = true
    skip_metabat2               = true
    skip_maxbin2                = true 

so I also would not expect any results from binners. depths/bin_summary.tsv doesnt exist in any case, I think.

Edit: clarity
Edit2: test_virus_identification run through, cannot find a problem, i.e. same output for nf-core/mag -r 3.1.0

OK that's strange, maybe I just had a weird test configuration (using the test_virus_identification profile, and extra flags), but it looks like you've tested in multiple configurations and it always looks OK 👍

@d4straub I will merge in and then do the template merge