Extending the Marker × Environment Interaction Model for Genomic‐Enabled Prediction and Genome‐Wide Association Analysis in Durum Wheat

Abstract

2193RESEARCH G enetic ´ environment interaction (G´E) affects trait heritability and the relative rankings of phenotypes across environments; this introduces challenges when making breeding decisions. The effects of G´E on heritability may be due to scale effects such as changes in the size of quantitative trait loci (QTL) effects across environments or to differential genetic effects on environmental variance. However, G´E can also modulate QTL effects, thus introducing changes in the relative rank of genotypes across environments (Dickerson, 1962;Cockerham, 1963). Falconer (1952) suggested modeling performance in two environments as two correlated traits; this allows modeling both scale effects and rerankings. Extending the Marker ´ Environment Interaction Model for Genomic-Enabled Prediction and Genome-Wide Association Analysis in Durum WheatJosé Crossa,* Gustavo de los Campos, Marco Maccaferri, Roberto Tuberosa, J. Burgueño, and Paulino Pérez-Rodríguez* ABSTRACTThe marker ´ environment interaction (M´E) genomic model can be used to generate predictions for untested individuals and identify genomic regions in which effects are stable across environments and others that show environmental specificity. The objectives of this study were (i) to extend the M´E model using priors that produce shrinkage and variable selection such as Bayesian ridge regression (BRR) and BayesB (BB), respectively, and (ii) to evaluate the genomic prediction accuracy of M´E, single-environment, and across-environment models using a multiparental durum wheat (Triticum turgidum L. spp. duram) population characterized for grain yield (GY), grain volume weight (GVW), 1000-kernel weight (GWT), and heading date (HD) in four environments. Breeding value predictions were generated for two prediction problems: cross-validation problem 1 (CV1) and cross-validation problem 2 (CV2). In general, results showed that the M´E model performed better than the single-environment and acrossenvironment models, in terms of minimizing the model residual variance, for both CV1 and CV2. The improved data-fitting gain over the other models was more evident for GWT and HD (up to twofold differences) than to GY and GVW, which showed more complex genetic bases and smaller single-marker effects. Considering the Bayesian models used, BB showed better overall prediction accuracy than BRR. As proofof-concept for the M´E model, the major controllers of HD-Ppd and FT on chromosomes 2A, 2B, and 7A-showed stable effects across environments as well as environment-specific effects. For GY, besides the regions on chromosomes 2B and 7A, additional chromosome regions with large marker effects were detected in all chromosome groups.

Genomic-Enabled Prediction in Maize Using Kernel Models with Genotype × Environment Interaction

“…This CV2 validation strategy performed better than CV strategy CV1 (where lines have not been evaluated in any field trials) proposed by Burgueño et al (2012), when applied in multi-environment models (Jarquín et al 2014; López-Cruz et al 2015; Crossa et al 2016b; Saint Pierre et al 2016). …”

Genomic assisted selection for enhancing line breeding: merging genomic and phenotypic selection in winter wheat breeding programs with preliminary yield trials

“…The prediction of individual trials or locations across years is an especially difficult task (Dawson et al 2013) and we observed a large variation in prediction accuracy for this undertaking in our study (Fig S3), fitting the results of other studies with autogamous crops (Heslot et al 2014; Lado et al 2016). Once multi-environment trials are being conducted, more options open up for enhancing the selection of variety parents like imputing untested lines in tested locations (Burgueño et al 2011; Jarquín et al 2014; Crossa et al 2016; Lopez-Cruz et al 2015) or enhancing the reliability of breeding values by a relationship matrix (Bauer et al 2006; Oakey et al 2007b; Bauer et al 2009; Müller et al 2015). Hence, predicting lines for the entire target population of environments might be a better strategy to select candidates that should enter multi-environment trials.…”

Bayesian Genomic Prediction with Genotype × Environment Interaction Kernel Models

“…The marker × environment interaction model has some advantages over previous models: it is easy to implement in standard software for GS and can be implemented with any Bayesian priors commonly used in GS, including not only shrinkage methods ( e.g. , GBLUP), but also variable selection methods (which cannot be directly implemented under the reaction norm model) (Crossa et al 2016). …”