This week we are excited to host one of the Vrije Universiteit’s very own researchers: Ronald de Vlaming. He discusses his research on the genetic architecture of human traits. There is a bulk of literature, in fields such as psychology and neuroscience, that shows genes help explain variation in human traits as well as correlations between traits, even in traits that are as far removed from biology as, say, educational attainment and risk aversion. That is to say, many human traits have a heritable component and are genetically correlated. In recent work, Ronald and his colleagues take research on estimation of heritability and genetic correlation a step further, by jointly fitting a model for a large number of traits. In the article found in the link below, he explains how genetic data can be used in statistical analyses in the first place and what kind of statistical analyses are needed to estimate heritability and genetic correlation. Next, he generalises this model to a multivariate setting and shows how matrix algebra and numerical methods can aid in rapid estimation of such a multivariate model.
If you find the application of techniques from Econometrics and Data Science in a field such as genetics to be of interest, we highly recommend Ronald’s article, which can be found under this link.