WDL-GWAS

Overview

With the UK Biobank RAP now serving as the exclusive access point to UK Biobank data, conducting large-scale statistical genetics analyses has become more complex. This repository offers a fully reproducible WDL workflow (RAP-compatible) for genome-wide association studies and fine-mapping, designed to run seamlessly on the RAP and adaptable for use in other computing environments.

"Workflow"

Installation

The workflow ca be run on linux platforms and macOS. However, for the later, you will need to change the local backend provider to run the workflow locally (see: Running WDL-GWAS Locally).

All dependencies require:

  • Java, which can be installed with sdkman.
  • Docker
  • This repository which you can obtain here.

Installation for Local Usage

Running the workflow locally requires cromwell, this page explains how to install it. Then, we recommend setting the environment variable CROMWELL_PATH to point to the newly downloaded jar file.

Installation for UK Biobank Usage

In order to run the WDL-GWAS workflow on the UK Biobank RAP you need to setup dxCompiler. This requires a few things:

  • Installing the dx toolkit.
  • Downloading dxCompiler from the releases page. Then, we recommend setting the environment variable DX_COMPILER_PATH to point to the newly downloaded jar file.

You can then login using your credentials via:

dx login