Skip to main content

November CIC Research Seminar

Date: November 13, 2019
Location: Innovation Central Perth - Building 216 Room 204 Level 2, Curtin University, Bentley Campus
CIC Seminar | Networking

Statistical challenges in identifying genetic component to depression

Presented by Prof Cathryn Lewis

Abstract:

The human genome comprises millions of genetic variants that are inherited within families. Many variants contribute to our risk of developing a disease, or play a role in continuous traits such as height or cholesterol level. Genetic data is a statistical goldmine and in the last decade, we have made great strides in uncovering the genetic predisposition to many disorders through genome-wide association studies. These studies have shown that most major public health problems are polygenic, with risk conferred by the combined effects of multiple genetic variants. Our most recent studies have identified 102 genetic variants, but these account for only a small proportion of trait variance, and of the genetic contribution.

Statistical challenges in these studies include:

1. Choice of relevant phenotype definition, for example, clinical diagnosis or number of depressive symptoms

2. Analysis strategies for millions of variants genome-wide using regression methods or machine learning methods.

3. Dissecting genetic correlations between disorders from shared variants

4. Prediction studies, from polygenic risk scores combining risk variants

In this talk, I will describe our studies to identify the genetic component to major depressive disorder. I will give an overview of the role of statistics in gene discovery, and highlight the opportunities in methodological and analytical statistical research based on the rich data arising from biological studies.

Bio:

Cathryn Lewis is Professor of Genetic Epidemiology and Statistics, with a split post across the Division of Genetics and Molecular Medicine (School of Medicine) and the MRC Social, Genetic and Developmental Psychiatry Centre (Institute of Psychiatry).

She leads King’s College London’s Statistical Genetics Unit, a cross-school group to develop and support methods in statistical genetics, with a particular focus on complex traits. She has held an academic post at King’s College London School of Medicine since 1996. Previously, she was at the University of Utah, Salt Lake City, and completed her PhD in Statistics at the University of Sheffield.

The one hour presentation will be followed by a networking opportunity.