Error executing process > 'PIPELINE_INITIALISATION:SAMTOOLS_FAIDX

[PasipanodyaL]

Description of the bug

Ran the pipeline successfully a few days ago then since yesterday it would not let me past the Samtools point.

Command used and terminal output

nextflow run nf-core/phaseimpute -r 1.0.0 -profile apptainer \
  --input /-/-/Allsamples_bootstap_samplesheet.fixed.csv \
  --posfile /-/-/phaseimpute_panelprep_clean/prep_panel/csv/posfile.quilt.csv \
  --chunks /-/-/phaseimpute_panelprep_clean/prep_panel/csv/chunks.csv \
  --steps impute --tools quilt \
  --fasta /-/-/GDDH13_v1_1/GDDH13_1-1_formatted.fasta \
  --outdir /-/-/phaseimpute_results \
  -w /scratch/user/-/phaseimpute_work_impute \
  -resume

Relevant files

No response

System information

No response

[LouisLeNezet]

Hi,
Thanks for the report.
Could you please provide a bit more details, like the nextflow output ?
When you say "samtools point" do you mean the SAMTOOLS_FAIDX process ?
If so it might be the region processing that fail.

[PasipanodyaL]

I was using conda I switched to apptainer then it worked
The full error message was: Error executing process > 'PIPELINE_INITIALISATION:SAMTOOLS_FAIDX (GDDH13_1-1_formatted.fasta)' Caused by: Process PIPELINE_INITIALISATION:SAMTOOLS_FAIDX (GDDH13_1-1_formatted.fasta) terminated with an error exit status (126) Command executed: samtools \ faidx \ GDDH13_1-1_formatted.fasta \ cat <<-END_VERSIONS > versions.yml "PIPELINE_INITIALISATION:SAMTOOLS_FAIDX": samtools: $(echo $(samtools --version 2>&1) | sed 's/^.samtools //; s/Using.$//') END_VERSIONS Command exit status: 126 Command output: (empty) Command error: .command.run: line 302: .command.run: Permission denied Work dir: /scratch/user/uqmpasip/phaseimpute_work_impute/9f/2d7c90f9f28e3ca61667d47ea00fcf Tip: view the complete command output by changing to the process work dir and entering the command cat .command.out

[LouisLeNezet]

This appears to be a permission issue with conda on your HPC filesystem. The error (Permission denied, exit status 126) typically occurs when conda environments are stored on filesystems mounted with noexec (a security restriction that prevents binary execution).

If you need to use conda, the fix would be to install conda environments on a local filesystem (not network-mounted storage) that allows execution:

// In your nextflow.config
conda {
    enabled = true
    cacheDir = '/tmp/conda_cache'  // Use local /tmp instead of /scratch
}

However if you are working on a HPC, I would highly recommend to use apptainer and add the following to your .bashrc, to tell nextflow and apptainer to pull all images in one location to save some space.

# Apptainer cache configuration for HPC
export NXF_APPTAINER_CACHEDIR ="/shared_full_path/apptainer"
export APPTAINER_CACHEDIR="/shared_full_path/apptainer"
export NXF_APPTAINER_ENABLED=true

Alternaly you can update your nextflow.config

apptainer {
    enabled = true
    autoMounts = true
    cacheDir = "/shared_full_path/apptainer"
}

or use your institution config with --profile my_institution_name.
If your institution is not yet present, I would recommend to ask your IT department to add it, as it could save space for everyone.

As it is solved on your side, I will close this issue, but feel free to reopen it if necessary or open a new now if you encounter any other issue.