Genomic selection: principles and possible uses in pig genetic improvement schemes

Genomic evaluation is a new methodology to estimate the breeding value of livestock. Its principle is to subdivide the genome into a very large number of chromosomal segments (CS) determined by one or several SNP (Single Nucleotide Polymorphism) markers, to estimate the effect of each of these CS on the traits of interest in a reference population (RP) phenotyped and genotyped for these markers, and then to calculate the genomic value of unphenotyped animals by summing the estimated effects of the CS carried by these individuals. The methods to estimate the effects of the CS are numerous, and vary in terms of accuracy and computing time. The efficiency of a selection based on this methodology increases with the heritability of the considered trait, the size of RP and the level of linkage disequilibrium between adjacent markers, but is lower in populations with a large effective size or when the genetic distance between the RP and the evaluated animals is large. In pigs, genomic selection could be used to improve populations for traits too costly to be recorded routinely, to increase the accuracy of the estimated breeding values, and to select the nucleus populations for their crossbred value expressed in a commercial environment. The implementation of genomic selection in pigs would be costly, mainly because of the size of the RP required to estimate accurate breeding values and the high number of candidates that should be genotyped. The sharing of RP by several breeding organizations and the use of genotypes imputation techniques would help reduce the implementation costs.