Integrative polygenic statistical approaches for genetic and epigenetic data to elucidate the origins of Alzheimer’s disease

• Applicant: Professorin Dr. Anke Hüls
• Subject Area: Epidemiology and Medical Biometry/Statistics
• Term: from 2018 to 2019
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 410552407

This project aims to extend and combine existing statistical approaches to improve the polygenic integrative analysis of genetic and epigenetic risk factors across multiple correlated phenotypes. Since Alzheimer’s disease (AD) is a polygenic disease with a complex phenotype that can be assessed by multidimensional cognitive tests, this project is based on data from two AD cohorts. Several genetic as well as epigenetic factors have been shown to be associated with AD. However, the mechanisms underlying these associations are generally unknown and the effect size for most variants is small resulting in a high proportion of missing heritability. Previous studies showed a phenotypic heterogeneity between different cognitive domains that can dilute any association between genotypes / DNA methylation changes and the global cognitive function. In addition, AD is known to be caused by polygenic risk factors, which clarifies the substantial interest in performing polygenic analyses to further elucidate the corresponding biological mechanisms behind disease development.The aims of this research project are 1) to identify genetic and epigenetic patterns that affect different cognitive domains measured with neuropsychological tests and 2) to evaluate the relative importance of epigenetic modifications on AD and its role for the association with depression and genotypes.In detail, in objective 1), the Gene Association with Multiple Traits (GAMuT) test will be extended to identify genetic and epigenetic patterns that affect different cognitive domains measured with neuropsychological tests. The GAMuT test was previously developed for rare-variant association testing of multiple phenotypes. In this project the GAMuT test will be extended to allow for i) methylation data and ii) multivariate ordinal phenotypes from neuropsychological tests, and iii) gene-network information that can potentially improve performance. In objective 2), polygenic risk scores will be constructed for i) epigenetic modifications (DNA methylation, called EGRS) and ii) genotypes (called GRS). SNPs/genes selection will be based on a priori knowledge from the literature and from the results of objective 1). These risk scores will be applied to elucidate the interplay between genetic and epigenetic factors, along with lifestyle choices and psychological factors. Interactions between GRS and EGRS will be investigated and mediation analysis will be conducted to investigate the mediating effect of EGRS on the associations between depression and AD and between GRS and AD.For evaluation and publication, all proposed statistical methods will be tested on simulated and real data sets. This project will improve the polygenic integrative analysis of genetic and epigenetic risk factors across multiple correlated phenotypes and further provide new insights of AD by incorporating genotypes, DNA methylation and psychological risk factors across multiple-domain cognitive test scores.

• DFG Programme: Research Fellowships
• International Connection: USA
• Host: Professor Michael P. Epstein, Ph.D.