From whole genome sequence to precision breeding

Genomic selection (GS) has to varying degrees been implemented in the breeding operations of the participating breeding organizations. However, due to its relative novelty, this technology is still rapidly improving nationally and internationally. Continuous R&D efforts are required from the participants to maintain at the forefront of this technology. The next step-change technology is the utilization of whole genome sequence (WGS) data, due to the dramatic cost reductions of whole genome sequencing, its overwhelming amount of information and, in particular, because causative mutations are included in WGS data. With this project, the partners prepare for the implementation of whole genome sequence data into their daily breeding operations. With respect to the operational tools for genomic prediction and selection the partners have a common goal, namely precision breeding. Precision breeding is defined by: (i) highly accurate genomic selection (>90%); (ii) selection does not deteriorate the functioning of the animals; and (iii) genetic variation is managed during the selection process, such that current selections do not jeopardize future opportunities for genetic gain. With elements (ii) and (iii) precision breeding advances the sustainability of animal breeding, because breeding that reduces the functioning of the animals and/or reduces genetic variation is not sustainable in the long term. This project will (1) focus genomic predictions towards the causative mutations; (2) generate WGS data; (3) optimize reference populations for the estimation of genetic effects; (4) manage genetic variation at the DNA level; and (5) test and implement the improvements into the breeding practices of the participating companies. The genome of 60 pigs has been sequenced and the data are in the process of being analysed using the new porcine reference genome. Currently another 180 pigs will be sequenced. Based on these whole genome sequences and relatively sparse genotyping of 50,000 Landrace pigs, the missing sequence data of the 50,000 Landrace pigs is being imputed. A data set for the robustness in the form of stayability of sows is being set together for genomic prediction. A list of bulls to be sequenced is generated, and is about to be sequenced. Methods have been developed that improved selection accuracy by 20% when improving salmon lice resistance. An international pig data set is in the process of being collected, which will demonstrate that the developed methods work across country, and elite breeding/practical herds boundaries. Computationally efficient large scale software has been developed that makes optimal use of whole genome sequence data for genomic prediction of breeding values of animals. The software has been tested on imputed whole genome sequence data of 35,000 bulls and cows. A method for correcting for missing pedigrees has been implemented such that simultaneously a correction for missing genotypes was implemented. For pigs and cattle, whole genome sequence data has been imputed based on sequenced individuals and densely genotyped individuals. The data has been combined with Dutch data (pigs) and Australian data (cattle) to perform across population genomic predictions. In cattle, the use of whole genome sequence data improved accuracies of genomic prediction by up till 3% compared to high density SNP chip data (600K). Across breed prediction increased accuracies by up till 1.5%. In pigs, similar improvements in accuracy of prediction were found.

De utviklede pipelines for hel genom-sekvensering og genotype og sekvensimputasjon implementeres av Norsvin og Geno. Hel genom -sekvensdatabaserte genomisk-seleksjon er ennå ikke implementert på grunn av kostnadene og fordi fordelene ble funnet å være moderate. De påviste og finkartlagde genene vil bli brukt til å hjelpe genomisk seleksjon i avlsorganisasjonene. Genomisk forvaltning av variasjon kombinert med genomisk seleksjon vil bli implementert av Norsvin og Geno, men mer forskning på effekter og detaljer av implementering i avlsprogram trengs.

At present, genomic selection (GS) has to varying degrees been implemented in the breeding operations of the participating breeding organizations. However, due to its relative novelty, this technology is still rapidly improving nationally and internationally. Continuous R&D efforts are required from the participants to maintain at the forefront of this technology. The next step-change technology is the utilization of whole genome sequence (WGS) data, due to the dramatic cost reductions of whole genome sequencing, its overwhelming amount of information and, in particular, because causative mutations are included in WGS data. With this project, the partners prepare for the implementation of whole genome sequence data into their daily breeding operations. With respect to the operational tools for genomic prediction and selection the partners have a common goal, namely precision breeding. Precision breeding is defined by: (i) highly accurate genomic selection (>90%); (ii) selection does not deteriorate the functioning of the animals; and (iii) genetic variation is managed during the selection process, such that current selections do not jeopardize future opportunities for genetic gain. With elements (ii) and (iii) precision breeding advances the sustainability of animal breeding, because breeding that reduces the functioning of the animals and/or reduces genetic variation is not sustainable in the long term. This project will (1) focus genomic predictions towards the causative mutations; (2) generate WGS data; (3) optimize reference populations for the estimation of genetic effects; (4) manage genetic variation at the DNA level; and (5) test and implement the improvements into the breeding practices of the participating companies. The project team was also responsible for the earlier introduction of genomic selection into the partners breeding schemes and thus has a track record of implemented research results.