The use of alternative genomic metrics in swine nucleus herds to manage the diversity of purebred and crossbred animals.

In livestock breeding populations, regions of the genome with a high frequency of runs of homozygosity (ROH) have reduced diversity. Metrics that reduce ROH frequency may provide an attractive way to manage the diversity and ensure long-term gains while avoiding inbreeding accumulation. The 2 objectives of the current work were to characterize the frequency of ROH in Landrace (LR), Large White (LW), and their cross (LR x LW) through a combination of real and simulated genotypes and to determine the impact of optimizing different inbreeding metrics for nucleus and crossbred populations. A ROH statistic (ROH5Mb: "1" if SNP was in ROH of 5 Mb and "0" otherwise) was calculated across the genome for genotyped LR (n = 1206) and LW (n = 1349) dams and simulated crossbred genotypes derived from mating 81 LR sires to 100 LW dams. High ROH5Mb frequencies were declared for a contiguous set of SNP within the top 5%. In addition to random mating, pedigree, genomic, or shared ROH-based relationship matrices were used to minimize relationships between mating pairs within breed and the latter 2 were in crossings. Mating plans with 25 sires available with a restriction on the maximum number of mating were devised for within-breed and across-breed mating populations of 625 and 1250 dams, respectively. Each plan was replicated 25 times. Regions of shared highROH5Mb frequency were found on SSC1, SSC3, and SSC14 and regions with a high ROH5Mb frequency within a breed were found to persist in the crossbreeds. Runs of homozygosity and genom-ic-based relationship matrices decreased the proportion of the overall genome in a ROH by 2.45- and 2.19-fold when compared with pedigree-based relationships. Furthermore, the use of pedigree-based relationships was not able to decrease regions with high ROH5Mb frequency more heavily than ROH-or G-based relationships. The use of alternative genomic re-latedness metrics such as ROH allow for relationships to be minimized for targeted regions of low diversity. [ABSTRACT FROM AUTHOR]