Hi again Mag team, I'm sorry for the multiple posts, I seem to keep running into walls. We are now running mag with the following command:

nextflow run nf-core/mag -r 5.3.0 -profile singularity --input samplesheet.csv --outdir /mag_test_01 --gtdb_db /DB/GTDB/226.0/auxillary_files/gtdbtk_package/full_package/release226 --skip_megahit --skip_prokka --gtdbtk_max_contamination 20

The first time we ran it without setting contamination threshold (allowing for the default at 10% max) however we did not receive any results for GTDBTk taxonomy classification. We compared a "failed" sample vs a "completed" one and noticed both had over 80% completeness. We thought maybe the had high contamination and were therefore not passing the gtdbtdk threshold. So we made the threshold more lenient. Even with the lenient threshold we are still not seeing taxonomy results. I would think we would at least see a gtdbtk_summary.tsv with failed samples, but that is not the case. Any direction would be appreciated.

Could you send a

.nextflow.log

file?

yes, here it is

OK looking at the log I see GTDBTK_CLASSIFYWF process for all samples completeing, but saying that many of the output files that get produced when GTDBTK does produce hits are missing

Jan-15 15:47:50.302 [TaskFinalizer-2] DEBUG nextflow.processor.TaskProcessor - Process NFCORE_MAG:MAG:GTDBTK:GTDBTK_CLASSIFYWF > Skipping output binding because one or more optional files are missing: fileoutparam<2:1>
Jan-15 15:47:50.302 [TaskFinalizer-2] DEBUG nextflow.processor.TaskProcessor - Process NFCORE_MAG:MAG:GTDBTK:GTDBTK_CLASSIFYWF > Skipping output binding because one or more optional files are missing: fileoutparam<3:1>
Jan-15 15:47:50.302 [TaskFinalizer-2] DEBUG nextflow.processor.TaskProcessor - Process NFCORE_MAG:MAG:GTDBTK:GTDBTK_CLASSIFYWF > Skipping output binding because one or more optional files are missing: fileoutparam<4:1>
Jan-15 15:47:50.302 [TaskFinalizer-2] DEBUG nextflow.processor.TaskProcessor - Process NFCORE_MAG:MAG:GTDBTK:GTDBTK_CLASSIFYWF > Skipping output binding because one or more optional files are missing: fileoutparam<5:1>
Jan-15 15:47:50.302 [TaskFinalizer-2] DEBUG nextflow.processor.TaskProcessor - Process NFCORE_MAG:MAG:GTDBTK:GTDBTK_CLASSIFYWF > Skipping output binding because one or more optional files are missing: fileoutparam<6:1>
Jan-15 15:47:50.302 [TaskFinalizer-2] DEBUG nextflow.processor.TaskProcessor - Process NFCORE_MAG:MAG:GTDBTK:GTDBTK_CLASSIFYWF > Skipping output binding because one or more optional files are missing: fileoutparam<7:1>

versus https://github.com/nf-core/mag/blob/ccfb744aab0c00a805a2dfbf7d9c9d8729c30805/modules/nf-core/gtdbtk/classifywf/main.nf#L17-L22 So I suspect in your results directory you'll find a log and

failed_genomes.tsv

. Is that correct? This implies GTDBTK cannot find anything that can be classified, thus does not generate the

*.summary.tsv

file that we use to produce teh

gtdbtk_summary.tsv

file. But it's hard to know this without seeing the GTDBTk logs. So I think the pipeline sort of working 'as intended' but suboptimally in that it doesn't inform the user that GTDBTk just doesn't find anything... If you could make an issue: https://github.com/nf-core/mag/issues We can see if we can improve this in the future 🙂