Our research is focused on the development and application of data-driven breeding techniques such as genomic selection for the identification of the most suitable parental combination, the use of optimal or near-optimal experimental designs for evaluating candidate varieties and the application of variation preserving selection strategies for balancing short- and long-term genetic gains. We investigate ways to integrate breeding technologies, from double haploid production to drone-based phenotyping, in data-driven breeding pipelines with the goal of establishing the most efficient roadmap towards genetic gain.
Quantitative traits are challenging to improve as they are controlled by numerous genes and influenced by varying environmental circumstances. Genomic selection has been rapidly adopted by plant and animal breeders to overcome these hurdles but many research paths remain unexplored and many questions remain unanswered. Why does genomic selection work well in some breeding programs while it fails miserably in others? What makes a good training population and what strategy should be used to improve an existing training population? How to incorporate the millions of predictors originating from next-generation DNA sequencing efforts while most genomic prediction models are already massively underdetermined? There is even more new ground to break in the related fields of genotype imputation and haplotype prediction, two components that are essential to the cost-effectiveness of genomic selection. We aim to be at the forefront of these scientific developments, forging a path towards a healthier and sustainable future for society and generations to come.
