AMPCOMBI2_PARSETABLES error during stage-out with process.stageOutMode = 'move' -- "mv: cannot stat 'XYZ': No such file or directory"

[amizeranschi]

Description of the bug

I am consistently getting errors from AMPCOMBI2_PARSETABLES when running on a cluster with process.stageOutMode = 'move'. The issue seems to occur during stage-out and results in error messages like the following:

ERROR ~ Error executing process > 'NFCORE_FUNCSCAN:FUNCSCAN:AMP:AMPCOMBI2_PARSETABLES (sample_2)'

Caused by:
  Process `NFCORE_FUNCSCAN:FUNCSCAN:AMP:AMPCOMBI2_PARSETABLES (sample_2)` terminated with an error exit status (1)


Command executed:

  ampcombi parse_tables \
      --path_list 'sample_2.ampir.tsv' 'sample_2.macrel.prediction' \
      --faa sample_2_pyrodigal.faa \
      --gbk sample_2_pyrodigal.gbk \
      --sample_list sample_2 \
      --amp_database DRAMP \
      --amp_database_dir amp_DRAMP_database \
       \
      --aminoacid_length 120 --db_evalue 5 --amp_cutoff 0.6 --ampir_file .ampir.tsv --amplify_file .amplify.tsv --macrel_file .macrel.prediction --hmmsearch_file .hmmer_hmmsearch.txt --ampgram_file '.tsv' --amptransformer_file '.txt' --neubi_file '.fasta' --log TRUE --hmm_evalue 0.06 --window_size_stop_codon 60 --window_size_transporter 11 \
      --threads 4
  
  cat <<-END_VERSIONS > versions.yml
  "NFCORE_FUNCSCAN:FUNCSCAN:AMP:AMPCOMBI2_PARSETABLES":
      ampcombi: $(ampcombi --version | sed 's/ampcombi //')
  END_VERSIONS

Command exit status:
  1

Command output:
  Success: 'ampcombi parse_tables' completed successfully ! Now run ampcombi complete !

Command error:
  mv: cannot stat 'amp_DRAMP_database/general_amps_2026_04_05.fasta': No such file or directory

Work dir:
  /data/share/users/amizeranschi/test-funcscan-error/work/66/f07f456eb57a447b5d07b451bb034e

Container:
  quay.io/biocontainers/ampcombi:2.0.1--pyhdfd78af_0

Tip: view the complete command output by changing to the process work dir and entering the command `cat .command.out`

 -- Check '.nextflow.log' file for details

Below is a minimal set of commands that should reproduce this error on a Slurm cluster, using the test profile:

cat << \EOF > custom.config
process.executor = 'slurm'
process.scratch = 'ram-disk'
process.stageInMode = 'copy'
process.stageOutMode = 'move'
process.errorStrategy = 'retry'
process.maxRetries = 2
EOF

rm -rf .nextflow* work test-ampcombi-error
clear
date
nextflow run nf-core/funcscan -r 60385fe \
-profile docker,test \
-c custom.config \
--run_amp_screening \
--amp_skip_amplify \
--run_arg_screening false \
--run_bgc_screening false \
--run_cazyme_screening false \
--run_taxa_classification false \
--run_protein_annotation false \
--outdir test-ampcombi-error

A local fix (which only deals with the symptom) is to set process.stageOutMode = 'copy' in the config. I have tested this and confirm that it "works". However, it would be more useful to fix this issue upstream in the module.

The error seems to be caused by overlapping outputs in the module, specifically here:
https://github.com/nf-core/funcscan/blob/dev/modules/nf-core/ampcombi2/parsetables/main.nf#L19

Both the main directory "${meta.id}/" and some of its content are declared as outputs. If process.stageOutMode = 'move' and the main directory is moved first, then Nextflow won't be able to subsequently stage-out the individual files as well, resulting in errors like the one above: mv: cannot stat 'XYZ': No such file or directory.

A similar issue is likely to occur with the "amp_${opt_amp_db}_database/" directory and its content:
https://github.com/nf-core/funcscan/blob/dev/modules/nf-core/ampcombi2/parsetables/main.nf#L26

Command used and terminal output

Relevant files

No response

System information

No response