Abstract

The sequencing of a number of the livestock genomes and developments in genotyping technology has made it possible to genotype animals for tens of thousands of markers relatively cheaply. Genomic selection exploits this information to calculate genomic breeding values (GEBVs), as the sum of the effects of small chromosome segments across the genome. Effects of chromosome segments are first estimated in a large population with phenotypic information. In subsequentgenerations, only marker information is required to calculate GEBVs. Accuracies of GEBVs calculated for animals with marker information only, can be as high as 0.85. Results from analysis of real data from dairy cattle suggest approximately 30,000 single nucleotide polymorphism markers will be required to calculate GEBVs with maximum accuracy. Chromosome segment effects should be estimated from at least 2000 phenotypic records. A trial of genomic selection in the Australian Holstein population, with 384 bulls typed for 9919 markers, achieved accuracies of GEBVs of up to 0.82, despite sub optimal marker density and limited number of phenotypic records. In dairy cattle, maximum genetic gain using genomic selection will be achieved by selection of young bulls for widespread use without progeny test, allowing a halving of the generation interval.

BJ, Hayes, and ME Goddard

Proceedings of the New Zealand Society of Animal Production, Volume 67, Wanaka, 143-146, 2007