Files
Abstract
The American pork industry experiences seasonal losses caused by heat stress. Genomic information can help to better identify heat-tolerant animals; however, heat stress evaluation requires complicated models. Single-step GBLUP (ssGBLUP) can be used for genomic selection with complex models and when only a fraction of the animals are genotyped. This method accounts for phenotype, pedigree, and genotype in a unified and simple approach; however, ssGBLUP has a limitation on the number of genotyped animals; it relies on direct inversion of the genomic relationship matrix (G); however, inverting a matrix has high computing cost, which creates a bottleneck. The number of genotyped animals is increasing at a fast rate for livestock species, and ssGBLUP would become unfeasible for more than 150,000 genotyped animals. The objective of the first study was to use genomic information to help mitigate problems associated with heat stress in the pork industry. Identifying a threshold for heat stress and including genomic information in the genetic evaluation increased the accuracy of prediction in production traits; therefore, ssGBLUP can be used to help mitigate the impact of heat stress on the US pork industry. The objective of the second study was to test a recursive algorithm, called algorithm for proven and young animals (APY), to compute the inverse of G in an efficient manner. In APY the genotyped population was divided into proven and young, and recursions were based on proven animals. In a simulated study with 25,000 genotyped animals, there was no significant difference between accuracy of GEBV obtained with regular or APY ssGBLUP, which indicate APY can successfully replace the direct inversion of G. A third study aimed to compare genomic predictions from regular and APY ssGBLUP for the US Holstein population; 100,000 genotyped animals were used in the study. Correlations of GEBV between the methods were greater than 0.99 when at least 10,000 animals were considered proven in the recursions. In general, genomic information can help mitigate problems due to heat stress in livestock species, and when the amount of genomic information is large, APY should be used in ssGBLUP to remove computing limitations.