Assessment of alternative genotyping strategies to maximize imputation accuracy at minimal cost

Pedigree and genotype data from a commercial pig population were used. Several alternative genotyping strategies were explored. The strategies differed in the density of genotypes used for the ancestors and the individuals to be imputed. Parents, grandparents, and other relatives that were not descendants, were genotyped at high-density, low-density, or extremely low-density, and associated costs and imputation accuracies were evaluated.Imputation accuracy and cost were influenced by the alternative genotyping strategies. Given the mating ratios and the numbers of offspring produced by males and females, an optimized low-cost genotyping strategy for a commercial pig population could involve genotyping male parents at high-density, female parents at low-density (e.g. 3000 SNP), and selection candidates at very low-density (384 SNP).Among the selection candidates, 95.5？% and 93.5？% of the genotype variation contained in the high-density SNP panels were recovered using a genotyping strategy that costs respectively, $24.74 and $20.58 per candidate.Successful breeding programs based on genomic information rely on large numbers of animals that are both phenotyped and genotyped at high-density [1,2]. Imputation of high-density genotypes for large numbers of phenotyped animals has been shown to be effective in generating large datasets at lower cost (e.g. [3-5]). Genotyping strategies for imputation generally involve genotyping some individuals in a pedigree at high-density, others at low-density, and in some cases not genotyping other individuals at all. Imputation of genotypes involves two steps. First, the haplotypes carried by the high-density genotyped individuals must be resolved. Then low-density genotypes are used in conjunction with pedigree, familial linkage, and linkage disequilibrium (LD) information to determine the combinations of haplotypes that are carried by animals that are not genotyped or that are genotyped at low-density. Several imputation algorithms h