ERROR: The flag 'directory' used in rule gtdbtk is only valid for outputs, not inputs.

[Maelinou61]

Hello,

I am trying to use MAGinator on the test data but it blocks indefinitely at this part :

[2025-05-13 14:28:06] INFO: Running MAGinator version 1.0.1 [2025-05-13 14:28:06] WARNING: Output directory results_test/ already exists [2025-05-13 14:28:06] INFO: Filtering bins [2025-05-13 14:33:18] INFO: 109 bins in 86 VAMB clusters left after filtering [2025-05-13 14:33:18] INFO: Classifying genomes with GTDB-tk

When I add the --only_conda argument to the maginator command I get this error :

[2025-05-15 11:09:44] INFO: Running MAGinator version 1.0.1 [2025-05-15 11:09:44] INFO: Filtering bins [2025-05-15 11:09:44] INFO: Only creating conda environments. [2025-05-15 11:09:51] INFO: Classifying genomes with GTDB-tk [2025-05-15 11:09:51] INFO: Only creating conda environments. [2025-05-15 11:09:53] INFO: Clustering genes and parsing GTDB-tk results [2025-05-15 11:09:53] INFO: Only creating conda environments. [2025-05-15 11:09:54] ERROR: Building DAG of jobs... [2025-05-15 11:09:54] ERROR: MissingInputException in rule parse_gtdbtk in file /exterieur/mrimeur/envs/maginator/lib/python3.12/site-packages/maginator/workflow/parse_g tdbtk.Snakefile, line 18: [2025-05-15 11:09:54] ERROR: Missing input files for rule parse_gtdbtk: [2025-05-15 11:09:54] ERROR: output: results_test/tabs/metagenomicspecies.tab, results_test/genes/all_genes.fna, results_test/phylo/intermediate/gtdb_markers.tab, result s_test/phylo/intermediate/gtdb_unique_bac_markers.tab, results_test/phylo/intermediate/gtdb_unique_ar_markers.tab, results_test/genes/all_genes.faa [2025-05-15 11:09:54] ERROR: affected files: [2025-05-15 11:09:54] ERROR: results_test/gtdbtk [2025-05-15 11:09:54] ERROR: Snakemake returncode: 1

It seems to be a snakemake error. I tried multiple snakemake version but it doesn't change anything. How can I solve this problem ?

Issues with the 1.0.2 version :

Hello,

I tried to use MAGinator with the 1.0.2 version. I followed again the installation steps but when I ran MAGinator with the test data and the --only_conda argument, I got this error :

[2025-05-22 12:10:06] INFO: Running MAGinator version 1.0.2 [2025-05-22 12:10:06] WARNING: Output directory test/ already exists [2025-05-22 12:10:06] INFO: Filtering bins [2025-05-22 12:10:06] INFO: Only creating conda environments. [2025-05-22 12:10:18] INFO: Classifying genomes with GTDB-tk [2025-05-22 12:10:18] INFO: Only creating conda environments. [2025-05-22 12:10:20] ERROR: The flag 'directory' used in rule gtdbtk is only valid for outputs, not inputs. [2025-05-22 12:10:20] ERROR: Building DAG of jobs... [2025-05-22 12:10:20] ERROR: MissingInputException in rule gtdbtk in file /exterieur/mrimeur/envs/MAGinator/lib/python3.12/site-packages/maginator/workflow/02_gtdbtk.Sna kefile, line 19: [2025-05-22 12:10:20] ERROR: Missing input files for rule gtdbtk: [2025-05-22 12:10:20] ERROR: output: test/gtdbtk, test/gtdbtk/gtdbtk.done [2025-05-22 12:10:20] ERROR: affected files: [2025-05-22 12:10:20] ERROR: test/combined_clusters [2025-05-22 12:10:20] ERROR: Snakemake returncode: 1

Moreover when I used MAGinator without the --only_conda argument then I had the same problem that I had with the previous version (indefinitely running).

It seems to be a snakemake error. I installed snakemake in different versions but I still have this problem.
So I am wondering which versions of Snakemake and Python do you use ?

Thank you!

Thank you!

[trinezac]

Hi, I've just updated the workflow as the GTDBtk-step seemed to sometimes stall.

I'm using Snakemake v8.28.0 and Python v3.12.9

Can you try to run it with --log_lvl DEBUG.

And maybe also send me the command you use.

Best Trine