Towards efficient and large routine genomic evaluations
Mario Calus is Senior researcher at the Animal Breeding and Genomics Centre. His main focus is on genomic selection and prediction, and within Theme 3 several people are working with him on genomic prediction for crossbred performance.
Mario Calus about Theme 3, (A)cross breed genomic selection:
The main aim of Breed4Food’s Theme 3 is to enable and optimize genomic selection of purebred selection candidates for the performance of their future crossbred offspring. This will provide an alternative to “conventional” selection that often selects purebred selection candidates based on purebred performance.
Whether genomic selection should be geared directly towards crossbred performance, or whether conventional selection is equally efficient, largely depends on the genetic correlation between purebred and crossbred performance. This correlation includes two components: 1) interaction of the genotype by the different environments in which the purebreds and crossbreds are kept, and 2) all “genetic” differences between purebred and crossbred animals. The second component is poorly understood, and therefore currently investigated by Yvonne Wientjes within the STW Breed4Food project “GenoMiX”.
Genomic prediction using crossbred information requires knowing the line-origin-of-alleles in crossbreds to link the performance of crossbred animals back to purebred selection candidates. Previously, John Bastiaansen, Claudia Sevillano and Jérémie Vandenplas focussed on deriving the line-origin-of-alleles in crossbred animals, and showed that we can derive this for >90% of the alleles. Currently, a genomic prediction model that uses line-specific relationship matrices is under investigation by Claudia Sevillano.
Once the industry starts to use crossbred information in their genomic evaluations, it is anticipated that the volume of this new data source will rapidly increase. To enable efficient computing for this and other scenarios, Jérémie Vandenplas started to implement the use of genomic recursions (also knowns as “the APY algorithm”) in the software calc_grm. This is part of his project that was funded through the 2015 Breed4Food call for bright ideas.
Results of the projects mentioned above will become available over the next few months, and will this summer be presented at scientific conferences. Once the APY algorithm and the use of line-origin-of-alleles in genomic prediction has been successfully tested, and shown its merit over and above currently used methods, these methods will be made available to all Breed4Food partners in a next release of the software calc_grm.
Do you have any questions about or suggestions for this study? Please contact Mario at mario.calus@wur.nl.
