Skip to content

Latest commit

 

History

History
105 lines (85 loc) · 3.88 KB

README.md

File metadata and controls

105 lines (85 loc) · 3.88 KB
Raw

Workflow

The workflows in this folder manage the execution of your analyses from beginning to end.

workflow/
 | - bin/                            Custom workflow scripts
 | - configs/                        Configuration files that govern workflow execution
 | - containers/                     Custom container definition files
 | - modules/                        Process definitions
 | - main.nf                         The primary analysis script
 \ - nextflow.config                 General Nextflow configuration

Usage:

Usage:

nextflow run -params-file <params.yml> [ -c <custom.config> ] [ -profile <profile> ] <nextflow script>

where:

  • params.yml is a YAML formatted file containing workflow parameters such as input paths to the data. See examples in the analyses subfolders. Alternatively parameters can be provided on the command-line using the --parameter notation (e.g., --samples <path> ).

  • <custom.config> is a nextflow configuration file which provides additional configuration or overrides existing settings. For example:

    // custom.config supplied using -c custom.config
process {
    withName: 'TASK' {
        cpus = 4       // Update cpus required
        time = 1.d     // Update maxium time process can run for
    }
}

  • <profile> is one of the preconfigured execution profiles (uppmax, singularity_local, docker_local). Alternatively, you can provide a custom configuration to configure this workflow to your execution environment. See Nextflow Configuration for more details.

Tool specific (module) parameters are supplied in the modules file. These can be overridden in the nextflow.config in the analysis directory, by using process selectors. For example:

// Module settings
process {
    withName: 'FASTQC' {
        // Override ext.args to FastQC
        ext.args = '--quiet'
    }
}

Workflow parameter inputs

Mandatory:

  • samples: A samplesheet containing sample information to analyse.

Optional:

  • results: The publishing path for results (default: results).

  • publish_mode: (values: 'symlink' (default), 'copy') The file publishing method from the intermediate results folders (see Table of publish modes).

    Software specific:

    • multiqc_config: Path to MultiQC configuration file (default: configs/multiqc_conf.yaml).

    Uppmax cluster specific:

    • project: NAISS Compute allocation number.

Workflow outputs

All results are published to the path assigned to the workflow parameter results.

  • 01_FastQC_sequence_check/: FastQC output
  • 02_Fastp_trimmed_sequences/: FastP output
  • pipeline_info/: A folder containing workflow execution details.
  • multiqc_report.html: A MultiQC report summarising the results of data processing.

Customisation for Uppmax

A custom profile named uppmax is available to run this workflow specifically on UPPMAX clusters. The process executor is slurm so jobs are submitted to the Slurm Queue Manager. All jobs submitted to slurm must have a project allocation. This is automatically added to the clusterOptions in the uppmax profile. All Uppmax clusters have node local disk space to do computations, and prevent heavy input/output over the network (which slows down the cluster for all). The path to this disk space is provided by the $SNIC_TMP variable, used by the process.scratch directive in the uppmax profile. Lastly the profile enables the use of Singularity so that all processes must be executed within Singularity containers. See nextflow.config for the profile specification.

The profile is enabled using the -profile parameter to nextflow:

nextflow run -profile uppmax <nextflow_script>