Nextflow Modules

UMCU Genetics Nextflow modules

Getting started with Nextflow

• Website: https://www.nextflow.io/
• Docs: https://www.nextflow.io/docs/latest/index.html
• DSL2 and Modules docs: https://www.nextflow.io/docs/edge/dsl2.html#modules
• Useful resources:
  • https://github.com/nextflow-io/patterns
  • https://github.com/oliverSI/GATK4_Best_Practice
  • https://github.com/SciLifeLab/Sarek

Nextflow workflows

• DxNextflowWorkflows

Contributing

We invite anybody to contribute to the UMCU Genetics Nextflow Modules repository. If you would like to contribute please use the following procedure:

• Fork the repository
• Commit the changes to your fork
• Submit a pull request

The pull request will be reviewed and included as fast as possible.

Coding guidelines

See utils/template.nf for a process template which uses the following guidelines.

• Use the Tool/version/Command.nf folder structure of this repository.
• Use the original tool version numbering
• Use CamelCase for tool, command and process names
• Use lowercase with words separated by underscores for params, inputs, outputs and scripts.
• Use 4 spaces per indentation level.
• All input and output identifiers should reflect their conceptual identity. Use informative names like unaligned_sequences, reference_genome, phylogeny, or aligned_sequences instead of foo_input, foo_file, result, input, output, and so forth.
• Define two labels for each process, containing toolname, version and command separated by an underscore.
  • BWA_0.7.17
  • BWA_0.7.17_MEM
• Define a tag to each process, containing toolname, command, sample_id and/or rg_id.
  • {"BWA MEM ${sample_id} - ${rg_id}"}
• Set a (hosted) container for each process.
• Add 'set -euo pipefail' to each process.
  • shell = ['/bin/bash', '-euo', 'pipefail']
• Do not define any runtime settings like cpus, memory and time.
• Set process parameters on include:
  • include process from 'path/to/process.nf' params(optional: '')
• Use separate process input channels as much as possible. Use tuples for linked inputs only.

  input:
        val(analysis_id)
        tuple(val(sample_id), path(bam), path(bai))
  

• Define named process output channels. This ensures that outputs can be referenced in external scope by their respective names. Indicate whether an output channel is optional.

  output:
        path("my_file.txt", emit: my_file)
        path("my_optional_file.txt",  optional: my_optional_file, emit: my_optional_file)
  

• Use params for resource files, for example genome.fasta, database.vcf.

GUIX

1. Creating squashfs immage

guixr pack -f squashfs -RR -S /bin=bin <name of tool you need> bash glibc-utf8-locales tzdata coreutils procps grep sed bootstrap-binaries

2. Copy .squashfs to appropriate directory and rename <Tool>_<version>.squashfs.
3. Add container to process: container = '<Tool>_<version>.squashfs'

Nextflow config for Utrecht HPC

profiles {
    sge {
        process {
            executor = 'sge'
            queue = 'all.q'
            errorStrategy = 'finish'
        }
    }
    slurm {
        process {
            executor = 'slurm'
            queue = 'cpu'
            errorStrategy = 'finish'
        }
    }
}

singularity {
    enabled = true
    runOptions = '-B /hpc:/hpc -B $TMPDIR:$TMPDIR'
    autoMounts = true
}