Skills Development Workshop

Day 1: Thursday, 12 September

This workshop will introduce participants to genetically informed causal inference methods and provide a practical example of how to run these anlayses. It will be split into three components: a 40 minute introduction to genetic causal inference methods; a 10 minute question and answer session; and a 40 minute practical on applying genetic causal inference methods. The talk will introduce participants to polygenic scores and Mendelian randomisation (MR). Polygenic scores are derived genetic measures that summarise known genetic predisposition to a trait such as BMI. We will discuss their conception and construction, how they may be biased by population level social factors, and how they can be derived or obtained from major research studies. We will also introduce MR, a methodological approach that uses genetic variants as instrumental variables to obtain unbiased (causal) estimates of health, behaviour and social exposures. We will discuss the assumptions of MR, implications when these are violated, when and how it can be applied, and how results should be interpreted. The practical will use a synthetic example dataset and open-source software, making use of Posit Cloud, an online version of RStudio. Using Posit Cloud will allow us to set up the data environment and scripts in a centralised location that participants can access. This approach therefore minimises risks typically encountered in practical sessions, such as participants not having the correct software/version installed, problems with data extraction, or incorrect data environments.