Important: This package is a KCL R interface to UKB data on Rosalind/CREATE. It is only useful for UKB-approved KCL reasearchers and their collaborators, with an account on the Rosalind/CREATE HPC cluster.
Overview
The ukbkings package includes functions to access UK Biobank (UKB) project data on the Rosalind/CREATE High Performance Computing (HPC) cluster, for UKB-approved King’s College London (KCL) researchers and collaborators. Access to data for a particular project is restricted to named collaborators on the project.
Installation
Install this development version from github with:
devtools::install_github("kenhanscombe/ukbkings", dependencies = TRUE, force = TRUE)Or, use the docker image to run R in a container with ukbkings and dependencies installed. Explicitly mount your study folder from your user directly for data storage, and the UKB data folder for access to UKB application data.
singularity run --bind /scratch/users/<username>/<study>:/<study>,/<path_to_datasets>/ukbiobank:/ukbiobank docker://onekenken/ukbkings:0.2.3Note: On Rosalind you first need to load the singularity module with module load apps/singularity/3.5.3; on CREATE the singularity command is immediately available.
Project data
Project-specific phenotype data and genetic link files (.fam, .sample, .dat) are in project-specific folders. The structure of a project directory setup for use with ukbkings is documented in the project setup tool, ukbproject.
Note. Any additional downloads (e.g. data baskets and associated keys, data downloads from the UKB data portal) should be stored in the subdirectory raw/. (Do not create further subdirectories within this folder)
