Nextflow Conventions

Nextflow Conventions

Single source of truth for Nextflow DSL 2 workflow standards. Other skills reference this. Follow nf-core guidance: https://nf-co.re/docs/tutorials/.

Copyright Boilerplate

All new .nf and .config files must start with:

// Copyright (c) 2026 Genome Research Ltd.
//
// Author: Sendu Bala 
//
// Permission is hereby granted, free of charge, to any person obtaining
// a copy of this software and associated documentation files (the
// "Software"), to deal in the Software without restriction, including
// without limitation the rights to use, copy, modify, merge, publish,
// distribute, sublicense, and/or sell copies of the Software, and to
// permit persons to whom the Software is furnished to do so, subject to
// the following conditions:
//
// The above copyright notice and this permission notice shall be included
// in all copies or substantial portions of the Software.
//
// THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
// EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
// MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
// IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY
// CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT,
// TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
// SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.

Project Layout

main.nf                     # entry workflow, includes subworkflows
nextflow.config             # default params + profile includes
conf/
  base.config               # default resource labels (process, memory, cpus, time)
  modules.config            # per-module publishDir and ext.args
  profiles.config           # profile definitions (docker, singularity, conda, test)
modules/
  local/                    # project-specific modules (one process per file)
    /
      main.nf
      meta.yml
      tests/
        main.nf.test
        nextflow.config     # test-specific config (ext.args etc.)
  nf-core/                  # installed nf-core modules (do not hand-edit)
subworkflows/
  local/                    # project-specific subworkflows
    .nf
workflows/
  .nf             # main workflow logic
lib/                        # Groovy helper classes/functions if needed
assets/                     # schemas, multiqc config etc.
docs/
  usage.md                  # end-user documentation: params, inputs, outputs
  output.md                 # description of all output files

DSL 2 Style

• nextflow.enable.dsl = 2 (implicit in modern Nextflow, but be explicit).
• One process per module file in modules/local//main.nf.
• Processes take tuple inputs with a meta map: tuple val(meta), path(reads).
• Emit named output channels: emit: bam, not positional.
• Use tag "${meta.id}" in every process for clear logging.
• Use label for resource classes: label 'process_low', 'process_medium',

'process_high'.

• Early check / error for required params at workflow entry, not inside

processes.

• Prefer channel operators (map, combine, join, groupTuple) over

Groovy logic in processes.

• No hardcoded paths. All input via params or channels.

nf-core Module Re-use

Before writing a local module, check if an nf-core module exists:

nf-core modules list remote | grep 
nf-core modules install 

Use nf-core modules where possible. Only create local modules when no suitable nf-core module exists. Do not hand-edit files under modules/nf-core/.

Containers

Every process must specify a container. Preference order:

1. nf-core module — container defined by the module already.
2. Seqera Wave / BioContainers — find at https://seqera.io/wave/ or

https://biocontainers.pro/. Prefer multi-tool BioContainers if available.

1. Docker Hub / Quay.io official images — for common tools.

Specify both docker and singularity in the process or config:

container "${ workflow.containerEngine == 'singularity'
    ? 'https://depot.galaxyproject.org/singularity/TOOL:VERSION'
    : 'biocontainers/TOOL:VERSION' }"

Software Versions

Every process must emit a versions channel. Use the eval output qualifier or capture version in the script block:

output:
tuple val(meta), path("*.bam"), emit: bam
path "versions.yml",            emit: versions

script:
"""
 --version 2>&1 | sed 's/.*//; s/ .*//' > versions.yml
cat  versions.yml
"${task.process}":
    : \$(echo \$( --version 2>&1) | sed 's/.*v//')
END_VERSIONS
"""

Collect all versions into a single versions.yml in the main workflow using CUSTOM_DUMPSOFTWAREVERSIONS or equivalent aggregation.

Configuration

nextflow.config — default params

params {
    input       = null
    outdir      = './results'
    // ... pipeline-specific defaults
}

// Include config files
includeConfig 'conf/base.config'
includeConfig 'conf/modules.config'

profiles {
    docker {
        docker.enabled = true
        singularity.enabled = false
    }
    singularity {
        singularity.enabled = true
        docker.enabled = false
    }
    conda {
        conda.enabled = true
    }
    test {
        includeConfig 'conf/test.config'
    }
}

conf/base.config — default resources

process {
    cpus   = { 1 * task.attempt }
    memory = { 6.GB * task.attempt }
    time   = { 4.h * task.attempt }

    errorStrategy = { task.exitStatus in [143,137,104,134,139,140] ? 'retry' : 'finish' }
    maxRetries    = 1
    maxErrors     = '-1'

    withLabel: 'process_low'    { cpus = 2;  memory = 12.GB; time = 4.h  }
    withLabel: 'process_medium' { cpus = 6;  memory = 36.GB; time = 8.h  }
    withLabel: 'process_high'   { cpus = 12; memory = 72.GB; time = 16.h }
}

Profiles

Profiles allow switching execution environments. Always support at minimum: docker, singularity, test. Institutional profiles can be added as needed.

Documentation

• docs/usage.md: describe all params, expected input format (e.g. samplesheet

CSV columns), how to run with examples.

• docs/output.md: describe every output file/directory produced.
• README.md: pipeline overview, quick start, link to docs.

Testing (nf-test)

• Framework: nf-test.
• Each local module has tests/main.nf.test beside its main.nf.
• Use snapshot assertions: assert snapshot(process.out).match().
• Wrap assertions in assertAll().
• Use minimal test data; for large data use -stub mode with stub blocks.
• Test both single-sample and multi-sample cases where applicable.
• Pipeline-level tests in tests/ at project root.
• Follow testing-principles for test intent; this section covers nf-test

mechanics.

• Every spec.md acceptance test MUST have a corresponding nf-test. No stubs

for logic tests, no hardcoded results, no swallowed failures.

nf-test file structure

nextflow_process {
    name "Test "
    script "../main.nf"
    process ""

    test("descriptive test name") {
        when {
            process {
                """
                input[0] = [
                    [ id:'test' ],
                    file(params.test_data['species']['type']['file'], checkIfExists: true)
                ]
                """
            }
        }
        then {
            assertAll(
                { assert process.success },
                { assert snapshot(process.out).match() }
            )
        }
    }
}

Commands

# Install nf-core modules
nf-core modules install 
nf-core modules list remote | grep 

# Run pipeline
nextflow run main.nf -profile test,docker --outdir results

# Run nf-test (all)
nf-test test

# Run nf-test (specific module)
nf-test test modules/local//tests/main.nf.test

# Lint
nf-core pipelines lint
nextflow run main.nf -profile test,docker -stub

# Clean work directory
nextflow clean -f

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: wtsi-hgi
• Source: wtsi-hgi/agentskills
• License: MIT

Install and usage instructions live in the source repository linked above.