I am working in the Breed4Food project in workpackage 4 "Genomic breeding program optimisation" (https://breed4food.com/). My goal is to develop methods that can potentially be used to improve any breeding program regardless of the species.
The two main focus areas are:
1) Improving selection. Traditionally, the animals with the best performance or with the best breeding value are selected as parents for the next generation. However, not the genetic value of the animal itself but its ability to produce selected offspring should be considered in parent selection. To select good parents, the Mendelian Sampling Variance needs to be taken into account. The aim of this topic is to develop and evaluate various criteria that consider MSV with regard to the ability to increase genetic gain.
2) Optimizing parameters of breeding programs. Breeding programs can be defined by many steps and actions that connect these steps (e.g. different selection intensities in a multi-stage selection program). To evaluate what values in each step (e.g. here: selection intensity) are best to bring the outcome of a breeding program as close as possible to its goal, a tool will be developed that gives the user recommendations what values should optimally be used for each parameter in the breeding program. This approachs does not require prior intuition for optimization and takes interdependency of parameters into account. This simplyfies the discovery of optimal settings.
The project is largely based on breeding program simulation. I am irregularly posting exemplary scripts on my GitHub (https://github.com/tobiasniehoff/exploring_MoBPS).
The techniques taht are developed in thsi project are validated and evaluated for contemporary commercial livestock breeding programs but can be applied to any other breeding program in principle.