Mendelian Randomisation

Mendelian randomization (MR) is a statistical technique that uses genetically characterised studies to understand cause-and-effect relationships among complex traits and diseases. Its wide use reflects the growing availability of genetic data and advancements in analytical tools and methods. Exciting developments in genetic data in terms of study designs, sample sizes, tissues genomically analysed and breadth of phenotyping are expanding the scope of causal inference questions, but these new data and questions must be served by statistical methods and tools that enable reliable causal inference. We will examine the complexities arising from rapid growth of genetic studies into millions of samples for MR, and determine how to use such resources, including family-based data, to most effectively improve causal inference. We will expand genetic studies by leading a new large international collaboration to systematically identify genetic influences that change across the lifecourse, and use such factors to identify time-critical disease development processes. We will develop new approaches to better model the molecular causes of disease, which will feed into our pharma collaborations for drug development. We will continue developing methods and data resources that expand epidemiological analysis of disease from prioritising causal factors of disease onset, to identify factors that influence prognosis and survival, creating the potential to identify novel therapies.