Adding modules to pipelines

nf-core modules are standardized, reusable components that wrap individual bioinformatics tools. Each module includes the tool’s process definition, container specifications, and metadata. By using modules from the central nf-core/modules repository, you benefit from community testing, standardized structure, and easier maintenance.

You can use nf-core tools to add these modules directly to your pipeline. Follow these steps to integrate a new nf-core module.

Add an nf-core module

To add an nf-core module to your pipeline with nf-core tools:

1. Install the module to your pipeline using the nf-core command-line tool:

   nf-core modules install <tool>/<subtool>

2. Add the include statement at the workflow level to import the module.

   • The installation command from the previous step will provide the exact syntax you need.

3. Insert the module call in the relevant workflow script.

   • Ensure version information is mixed into the versions channel and route output files to MultiQC if applicable.

4. Create a dedicated section in conf/modules.conf that specifies the default output directory and file patterns for the module.

5. Add module-specific parameters in nextflow.config with sensible defaults.

   • For pipeline-level parameters that need to be passed to the module, insert them into ext.args using proper closure syntax:

     ext.args = { "--option $params.option" }

6. Update nextflow_schema.json to include the new parameters:

   nf-core pipelines schema build

7. Update the following documentation files:

   • Add tool citations to CITATIONS.md
   • Document usage information in docs/usage.md
   • Describe output directories in docs/output.md
   • Update the pipeline README and workflow diagrams
   • Update the MultiQC configuration file

8. Run the nf-core linting tool to ensure your code meets nf-core standards:

   nf-core pipelines lint

9. Run Prettier to format your code.

   prettier -w

10. Run the pipeline locally with test data to verify:

    • The module executes successfully
    • Outputs are generated correctly
    • No errors or warnings occur