Effect of genotyping strategies in the sustained genetic gain across multiple generations of selection using ssGBLUP

Single-step GBLUP (ssGBLUP) is a BLUP-type genomic evaluation method, which uses a relationship matrix calculated using genotype and pedigree information, allowing to include both genotyped and ungenotyped individuals into a single evaluation. The practical implication of this method is that the genotype information propagates across all animals, improving the accuracy of the breeding values of ungenotyped candidates. A simulation study was carried out to quantify the impact of the genotyping strategy on the sustained genetic gain across multiple generations. To mimic a sheep population, the population structure was simulated with large paternal half sib and small maternal half sib
families, but discrete generations were also assumed to avoid changes in generation interval affecting the results. Each generation was composed of 900 individuals (300 males, 600 females) from which 30 males were selected and, each mated to 10 selected females (300 females in total). To create an initial reference population, the first 2 generations were unselected with all animals being genotyped (i.e., initial population of size 1800). Thereafter, a proportion of the candidate offspring were genotyped and the genetic evaluation carried out using ssGBLUP including all candidates plus all ancestor animals since the first generation. The scenarios compared were standard BLUP, GBLUP with all genotyped animals and 5 ssGBLUP with the number of candidates genotyped being 10%, 20%, 30%, 40% and 50%, respectively. The strategy to select the proportion of genotyped animals were: random and the top best animals based on their phenotypic record. The selection scheme was carried out for 9 generations (11 generations simulated). Results showed that the genetic gain of ssGBLUP increases almost linearly with the proportion of individuals genotyped. The comparison between genotyping strategies showed that the selection at random yielded greater genetic gain than when genotyping the best animals in the first generation. However, in the following generations, the scheme where the best animals are genotyped yielded grater cumulative response.