Gene-by-environment interactions in depression

Background
Major depression (MD) is one of the commonest psychiatric illnesses and one of the world's leading causes of morbidity. It is a heritable condition, with a heritability estimated to be between 30 and 40%. The number of associations detected by genome-wide association studies (GWAS) is still low but is gradually increasing with larger sample sizes. MD also has an important environmental component, but the effect and scale of interactions between genes and environment has been rarely studied. This project will initiate the identification of these interactions using computational analysis of data from Generation Scotland (GS), UK Biobank (UKB) and the Swedish GAPs (GAPs) project.
GS is a family-based population study comprising data from more than 20,000 individuals from across Scotland. UKB is a health resource consisting in 500,000 individuals from all over Britain. In both cohorts records from several health-related complex traits (including depression status) are available, together with lifestyle ("environmental") data (e.g., diet, income, etc.). In addition, genotypes and imputed data are available in both cohorts.
Replication and extension will be possible using available samples from Sweden. The GAPS project is accruing genomic data and register-based phenotypes (medical treatment for MDD, chronic antidepressant use) for over 150,000 subjects including 10,000 MDD cases.

Aims
We will use several complementary computational approaches to explore gene by environment interactions (GxE) impacting clinical and self-declared depression and correlated traits such as neuroticism. Firstly we will use a genome-wide and composite environment data (combining several lifestyle variables together), using computational analysis to estimate the overall amount of trait variance explained by GxE. Secondly, GxE will be explored from an individual locus and specific environment perspective. Here we will determine how the effect of a locus on a trait depends on individual lifestyle choices and other environmental data (e.g., diet). This approach will also be extended to a group of relevant genes (for example applying polygenic risk scores based on GWAS that have identified a number of SNPs associated with depression or correlated traits). The individual GWAS SNPs will also be tested for interactions with different environmental variables in order to identify interactions between specific loci and environments. Methylation data as proxy environmental data will also be incorporated into statistical analyses.
The objectives of this project are: (1) to gauge the extent of variation accounted for by GxE in depression and related traits in Generation Scotland, UK Biobank and GAPS and (2) to dissect GxE further by through interactions between specific loci and environments.
GxE has been of major interest to researchers in depression for many years, and to researchers of resilience, who seek to identify why some people appear more vulnerable to adversity than others. It is very unusual, however, to have the sample size, study design or methods available to identify the presence of GxE with sufficient confidence to allow potentially illuminating new insights. These would improve disease risk prediction and help stratify patients for better prognosis. The current project brings all of these ingredients together in a PhD project that is challenging, but may reveal findings of significant importance to the fields of depression and resilience research.

Training outcomes
- Large scale computational genetic data analyses (variance component analyses, association analyses, heritability estimation)
- Computational skills (R programming, BASH and potentially other scripting languages)
- Use of major computational packages (PLINK, GenABEL...)
- The biology, diagnosis and treatment of MD
- Analyses of questionnaire data
- Construction of genetic risk scores
- Research in an international collaboration