Correction: Impact of dominance effects on autotetraploid genomic prediction

Many commercially important plants are autopolyploid. As a result of the multiple chromosome sets in their genomes, higher orders of allele interactions can occur, implying different degrees of dominance. In contrast with diploids, dominance effects can be heritable in polyploids, potentially having a higher impact on the prediction of genetic values. In this study, we investigated the role of additive and dominance effects in the prediction of genotypic values for complex traits in autotetraploid species in the context of genomic selection. As autotetraploid model species, we used data from breeding populations of blueberry (Vaccinium spp., n = 1804) and potato (Solanum tuberosum L., n = 560), assessing genetic parameters and prediction ability of five and two horticultural traits, respectively. Using a Bayesian framework, the genotypic effects were estimated based on (i) realized additive and digenic dominance relationship matrices, and (ii) all markers included as explanatory variables under ridge regression and Bayes B approaches. When included, dominance effects explained part of the estimated genetic variance and resulted in better goodness-of-fit values. However, their predictive ability was similar to the predictability obtained with additive models. Although we have considered only autotetraploid species in this study, many of the ideas and results should be of more general interest, with applications in species with higher ploidy level. R.R. Amadeu, L.F.V. Ferrão, I.D.B. Oliveira, J. Benevenuto, and P.R. Munoz, Blueberry Breeding and Genomics Lab, Horticultural Sciences Dep., Univ. of Florida, Gainesville, FL 32611; J.B. Endelman, Dep. of Horticulture, Univ. of Wisconsin, Madison, WI 53706. Received 28 Feb. 2019. Accepted 22 May 2019. *Corresponding author ([email protected]). Assigned to Associate Editor Carlos Messina. Abbreviations: A-BayesB, additive Bayes B regression; A-BLUP, additive best linear unbiased prediction; A-BRR, additive Bayesian ridge regression; A+D-BayesB, additive + dominance Bayes B regression; A+D-BLUP, additive + dominance best linear unbiased prediction; A+D-BRR, additive + dominance Bayesian ridge regression; BayesB, Bayes B regression; BRR, Bayesian ridge regression; DIC, deviance information criteria; GBLUP, genomic best linear unbiased prediction model; GEBV, genomic estimated breeding value; Gnr-BayesB, general Bayes B regression; Gnr-BRR, general Bayesian ridge regression; GS, genomic selection; ls-means, least squares means; PA, prediction accuracy; P-BLUP, pedigree-based best linear unbiased prediction; PCA, principal component analysis; SNP, single nucleotide polymorphism. Published in Crop Sci. doi: 10.2135/cropsci2019.02.0138 © 2019 The Author(s). This is an open access article distributed under the CC BY license (https://creativecommons.org/licenses/by/4.0/). Published online August 1, 2019