Research overview
Research projects in this theme involve the use of Mendelian randomization (MR), genetic colocalization and a range of biological/biomedical databases to predict the efficacy and safety of drug targets. Work in this theme involves collaboration with several pharmaceutical companies.
Off-target side-effects
A key challenge in the use of MR for drug target prioritisation is that the approach typically only predicts on-target side-effects. In work led by alumnus and collaborator Jie Zheng, we have used MR to evaluate the effects of multiple targets of metformin. We are extending this to other multi-target drugs, and using new datasets to identify undocumented alternative targets so that we can more effectively predict off-target effects.
Integrating drug target evidence
In work led by Senior Research Associate Maria Sobczyk-Barad, we integrated data from a range of sources to improve interpretation of hits from genome-wide association studies (implemented in the MendelVar tool). We will continue to build on this concept by integrating different evidence sources with MR evidence from multiple molecular traits (gene expression, protein levels, DNA methylation) to better predict efficacy and safety.
Cross-ancestry Mendelian randomization
Genome-wide association studies have historically been focused on European-ancestry populations, with the result that most MR studies are better powered to detect effects in these populations. In collaboration with the Global Biobank Meta-analysis Initiative (GBMI), Jie Zheng and PhD student Huiling Zhao carried out a cross-ancestry proteome-wide MR analysis to identify drug targets across different populations. In a cross-MRC unit collaborative network led by Senior Research Associate Amanda Chong, we are now working on new analyses of potential drug targets across different populations. (Work in this area is in collaboration with Gibran Hemani and the IEU Mendelian randomization programme).
Tissue/pathway-based Mendelian randomization
Genetic variants associated with complex traits such as obesity represent a range of underlying molecular pathways, with potentially differential effects on obesity-linked diseases. Identifying these pathways could identify more specific intervention targets that block the pathway from obesity to disease. In work led by Maria Sobczyk-Barad, we built on previous research by Leyden et al to show distinct pathway-specific effects of blood pressure and adiposity on cardiovascular traits. In related work, Senior Research Associate Hayley Wragg is working on new clustering approaches for genotype-phenotype associations to help identify potential pathways. (Work in this area is in collaboration with Gibran Hemani and the IEU Mendelian randomization programme).