Introduction

RAFT provides a consistent, reproducible way to run bioinformatics analyses across samples, environments, and workflow versions. Whether you are processing a single patient or a large cohort, every RAFT project follows the same structure — from setup through results.

What RAFT does

  • Manages projects — initializes a standard layout and tracks what ran, when, and with what inputs

  • Runs versioned workflows — pulls specific workflow versions from GitLab and executes them through Nextflow

  • Stages inputs — helps you build sample manifests, download public datasets, and wire inputs together before execution

  • Surfaces results — launches a browser-based viewer for QC metrics, reports, and workflow outputs

  • Works anywhere — runs on a local cluster (SLURM and others) or your own GCP infrastructure

How a RAFT project works

RAFT projects all follow the same basic flow:

  1. Set up your workspace

  2. Choose a workflow

  3. Prepare inputs

  4. Run the analysis

  5. Review results

Where to start

New to RAFT?

Follow Quick start to install RAFT and run a demo in minutes.

Ready to run your own data?

Use Choosing your environment to decide between local and cloud, then follow the relevant setup guide.

Looking for a specific command or option?

See Running RAFT locally, Running RAFT on the cloud, and Manifest Specifications for execution guidance and manifest details.