Skip to navigation Skip to main content Skip to footer

Approved research

Comprehensive Phenotype-wide Association Studies (PheWAS)for Genetic Tool Variants relevant to GSK Drug Targets

Principal Investigator: Professor John Whittaker
Approved Research ID: 20361
Approval date: June 13th 2016 | Completion date: September 20th 2018

Lay summary

Overall success rates for bringing novel medicines to patients are low. Reasons for failure in drug discovery and clinical development are many and complex, including choosing wrong target-indication pair(s) and limited understanding of the biology and mechanisms of action. It is now widely accepted that genetic associations with disease phenotypes may constitute ?drug target validation? evidence, with improved likelihood of success. This study aims to perform systematic Phenome-wide association studies (PheWAS) to evaluate associations between relevant drug-target genes and all health-related outcomes. All genotyping, health history, biochemistry and linked health-related outcomes will be requested. This research will systematically evaluate associations between relevant drug target gene variants and all health-related outcomes. This research will provide useful results to validate existing target-indication pairs, and discover alternative indications for existing drugs. Adding human target evidence in portfolio progression decisions may increase success rates in subsequent clinical development. Focusing investment in drugs/targets most likely to ultimately deliver patient benefit reduces patient numbers enrolled in trials that will ultimately fail (wasting patient volunteer effort and associated risks to patients). This fulfills the UK Biobank?s stated purpose to improve the prevention, diagnosis and treatment of illnesses. Variants and combinations of variants in existing and potential drug target genes will be evaluated for association with all health-related and disease outcomes as well as markers of disease severity and progression recorded within the UK Biobank. Initially, analyses will be performed using simple groupings of ICD-10 codes, followed by more detailed evaluations of disease outcomes and necessary sub-phenotypes. Genetic association analyses will utilize regression approaches, adjusting for potential confounders. Results will be submitted for publication in peer reviewed journals. We will request data for the full cohort