Statistical Genetics

Natural variation is the prime resource to discover the evolutionary changes associated with adaptation to changing environmental conditions. To retrieve the genetic variants important for adaptation, genome-wide association mapping (GWAS) has become the standard technique. This method was pioneered in human genetics ten years ago and is since routinely used in many organisms. Additionally the standard model for GWAS has been extensively improved: State-of the art methods are nowadays super-fast, can detect allelic heterogeneity and incorporate gene-environment interactions. Building on these models, it is possible to identify and understand the natural variation in molecular signalling pathways. I am interested in both the development of new statistical tools for association mapping and the identification of causal relationships between genotypes and phenotypes. For this reason I work mainly with data from the model plant Arabidopsis thaliana. The benefits of Arabidopsis for GWAS have been already highlighted and it is the perfect model for the questions I want to ask. Additionally, the amount of available data is incomparable.

In more detail, I'm really interested to shift the analysis of natural variation to a cellular level, which again is possible using Arabidopsis.

The phenotype I care most about is drought tolerance, as water availability has a major impact for plant growth and development and is of agricultural relevance. Here, naturally evolved variants are prime targets for crop improvement.  

GWAS method development for more sophisticated GWAS methods

* including epistasis in GWAS setting

* Haplotype-based methods

* GWAS on GPUs