[Do not merge!] Pseudo PR for first release

[mashehu]

Do not merge! This is a PR of dev compared to the TEMPLATE branch for whole-pipeline reviewing purposes. Changes should be made to dev and this PR should not be merged!

[mashehu]

I think I went through everything now.

Good job on writing this pipeline, not an easy one to do.
I left several comments that apply in similar cases as well, please always fix them in all of them.

One major thing that needs to be added imo is some kind of parameter.csv validation. Based on the docs you have already quite a good knowledge of what types and what are the possible values etc. I would write these into a JSON-schema (LLMs are quite good at that) and then have a process like this to validate the csv files:

process VALIDATE_PARAMETERS {
    tag "$meta.id"
    
    input:
    tuple val(meta), path(parameters_csv)
    path schema_json
    
    output:
    tuple val(meta), path(parameters_csv), emit: validated
    
    script:
    """
    #!/usr/bin/env python3
    import json
    import csv
    import sys
    from jsonschema import validate, ValidationError
    
    # Load schema
    with open('${schema_json}') as f:
        schema = json.load(f)
    
    # Read CSV and convert to dict/list for validation
    with open('${parameters_csv}') as f:
        reader = csv.DictReader(f)
        data = list(reader)
    
    # Validate
    try:
        validate(instance=data, schema=schema)
        print(f"✓ Validation passed for ${meta.id}")
    except ValidationError as e:
        print(f"✗ Validation failed for ${meta.id}: {e.message}")
        sys.exit(1)
    """
}

[mashehu]

is this really needed?

[CaroAMN]

i get a linting error if i don't have this .... i didn't know to solve it otherwise

[mashehu]

this feels like a generic module and not a tool. does it really generate output in the results folder?

[CaroAMN]

Yes, for some output files that are in .mat format, it converts them into JSON or CSV. Because .mat files can only be opened by MATLAB

[mashehu]

same as before. for subworkflows caputring the process outputs is preferred

[CaroAMN]

i can change that , but i saw in the documentation now that for subworkflows one should check with workflow. and i think it would make sense to do so.
https://nf-co.re/docs/tutorials/tests_and_test_data/components/04_testing_subworkflows

therefore i think i would leave it like that. Or maybe i misunderstood something