Background

The Nextflow workflow manager is a powerful tool for scientific workflows. In order for end users to launch a given workflow with different input data and varying settings, pipelines are developed using a special variable type called parameters (params). Defaults are hardcoded into scripts and config files but can be overwritten by user config files and command-line flags (see the Nextflow docs).

In addition to config params, a common best-practice for pipelines is to use a "sample sheet" file containing required input information. For example: a sample identifier, filenames and other sample-level metadata.

Nextflow itself does not provide functionality to validate config parameters or parsed sample sheets. To bridge this gap, we developed code within the nf-core community to allow pipelines to work with a standard nextflow_schema.json file, written using the JSON Schema format. The file allows strict typing of parameter variables and inclusion of validation rules.

The nf-schema plugin moves this code out of the nf-core template into a stand-alone package, to make it easier to use for the wider Nextflow community. It also incorporates a number of new features, such as native Groovy sample sheet validation.

Earlier versions of the plugin can be found in the nf-validation repository and can still be used in the pipeline. However the nf-validation plugin is no longer supported and all development has been moved to nf-schema.