Mendelian randomization: George Davey Smith
From Mendelian randomization to hypothesis-free causal inference.
This innovative programme aims to extend the now widely utilized Mendelian Randomization (MR) paradigm in ways that will substantially increase its utility in population and clinical sciences.
Aims and objectives
The primary objective of this research is to develop MR to:
- Address the recognized limitations of MR through development of a range of approaches, to address issues such as limited statistical power, pleiotropy and the lack of established genetic instruments
- Incorporate increasingly accessible omics data (epigenetic and metabolomic) to assist in defining the role of these intermediate pathways in disease causation
- Harness vast wealth of genome-wide SNP data for application in MR approaches, increasing the power and scope of this widely applicable tool for causal inference.
Linkage to the other MRC IEU programmes will be integral to the methodological development and application of MR. The integration of epigenetic data and the development of two-step epigenetic MR will be achieved in association with programme 2. Two sample MR will be utilized in programme 3. Pleiotropy Robust Mendelian randomization will involve the exploitation of GWAS data in MR approaches as well as contributing to GWAS of DNA methylation to underpin two-step epigenetic MR. Programmes 5 and 6 will apply novel MR developments to cardiometabolic health and health-related behaviours.