Additive and heterotic effects estimation from a F2 Duroc × Pietrain crossbreed using 60K realized breed composition and heterozygosis

ROGBERG MUÑOZ A; FORNERIS NS; STEIBEL JP; MUNILLA LEGUIZAMÓN S; ERNST CW; BATES RO; GIOVAMBATTISTA G; CANTET RJC

Salt Lake City

Conferencia; 35th International Society for Animal Genetics Conference; 2016

International Society for Animal Genetics

Heterosis is mainly the result of non-additive geneticeffects. It is defined as the difference between theaverage phenotype of the F1 progeny minus the averageof the parental lines. If heterosis is caused bydominance, the values of heterosis are proportionalto the loss of heterozygosity, which in turn is definedas the average fraction of the genome with heterozygousloci from the parental lines. Assuming a 2-locusmodel of genetic effects, the expected value of thephenotype in a F2 progeny from 2 breeds (C and T)can be written as follows: E(CT) = μ + kAA + kDD +kAAAA + kDDDD + kADAD. The coefficients kA and kDare functions of the genomic proportion of each breed(S), and of the fraction of the genome in heterozygousstate (H), respectively. Usually S and H have been?estimated,? using individual breed composition suchthat all animals at the same crossbreeding stage havethe same values of S and H. With the developmentof SNP arrays technology, thousands of positions inthe genome could be genotyped and inheritance ofgenomic segments could be precisely estimated. Thus,S and H can be ?observed? in a specific cross withan adequate method of genotyping. The goal of thisresearch was to estimate the additive and dominanceeffects from a Duroc × Pietrain F2 crossbred population,using phenotypic data from the F2 animals andgenomic data from the F0, F1, and F2. A total of 411pigs from an outbred 3-generation Duroc × Pietrainresource population were genotyped with the PorcineSNP60Beadchip. Data on growth (15 traits) andcarcass (25 traits) were also collected from the F2 generation(336 animals). For any pair of grandparent?grandson, the identity by descent sharing within theknown pedigree, across all SNPs, were inferred withthe software PEDIBD and realized values of S and Hwere estimated in the grandsons by using as referencethe breeds of each grandparent. Then, the estimatedvalues of S and H were used to calculate kA and kDfor each F2 animal, values that were later included ascovariates in a mixed model for estimating the additive and heterotic effects. The realized values of kA and kDranged between 0.480 to 0.377, and from 0.716 to0.360, respectively, and allowed estimating A, D, AA,AD, and DD for all 40 traits.