The contribution of de novo mutations to long-term selection response in genomic breeding programs (Mutabreed)

The amount of data that is accumulating over the lifespan of animals is increasing very rapidly, but the technology to analyse it to full potential is lagging behind. We propose to investigate the applicability of advanced machine learning methods to get insight in the variation in phenotypic patterns over time, and in the value of adding data on the life history of an animal to the genomic information to improve the prediction of the animal?s future phenotype. We will use the 2classification method called Generalized Matrix Learning Vector Quantization (GMLVQ) to predict probabilities that the performance of an animal will fall into a predefined set of classes. We will also investigate ensemble methods that combine the results of different predictive models. In particular, we will investigate the modelling of GMLVQ using genotype and environment type of data and multiple decision trees (referred as random forests) for the modelling of various types of phenotype data. Finally, the output of the GMLVQ and the decision trees will be fed to another classifier (GMLVQ or decision tree or other) that computes the performance probability of a given animal. The expected main results are: 1. A method that identifies animals with performance that (temporarily) deviates from the expectation. This would allow for early prediction and identification of reduced health so that timely action can be taken. This would allow selection for increased health status based on data that can be standardly recorded on large numbers of animals at no additional costs; 2. A method to optimise use of large data, including genomic and time serial phenotypical, and environmental data, for predicting future performance of an animal. This would allow for pre-sorting of animals, for example finisher pigs to groups targeted at intended market, but also for increased accuracy of predicted phenotypes that will be available prior to the actual phenotypic performance, for example whether or not a cow will reach her second lactation.

  • Projectnummer / Project number: 14297
  • Deelnemende kennisinstellingen / Participating institutes: Wageningen University & Research - University of Groningen
  • Projectleider / Project leader: dr. ir. H.A. Mulder - Wageningen University & Research - Dierwetenschappen - Animal Breeding and Genomics Centre
  • Type project / Type project: Initieel
  • Startdatum / Start date: 01-07-2018
  • Programma / Programme: STW-Breed4Food


Lisette Krul, MSc  (Programmamedewerker) Lisette Krul, MSc (Programmamedewerker) +31 (0)30 6001305