Genomic selection using random regressions on known and latent environmental covariates

• DOI: 10.1007/s00122-022-04186-w
• Publication Date: 2022-09-06T05:03:41Z
• Authors: Daniel J. Tolhurst; R. Chris Gaynor; Brian Gardunia; John M. Hickey; Gregor Gorjanc
• Keywords: Plant biochemistry; Genomic Selection

Abstract Key message The integration of known and latent environmental covariates within a single-stage genomic selection approach provides breeders with an informative practical framework to utilise genotype by environment interaction for prediction into current future environments. This paper develops which integrates special factor analytic framework. linear mixed model Smith et al. (2001) is effective method analysing multi-environment trial (MET) datasets, but has limited practicality since the underlying factors are so modelled (GEI) observable, rather than predictable. advantage using random regressions on covariates, such as soil moisture daily temperature, that GEI becomes integrated (IFA-LMM) developed in this includes predictable observable terms joint set covariates. IFA-LMM demonstrated late-stage cotton breeding MET dataset from Bayer CropScience. results show predominately capture crossover explain 34.4% overall genetic variance. most notable maximum downward solar radiation (10.1%), average cloud cover (4.5%) temperature (4.0%). non-crossover 40.5% also accuracy $$0.02-0.10$$ <mml:math xmlns:mml="http://www.w3.org/1998/Math/MathML"> <mml:mrow> <mml:mn>0.02</mml:mn> <mml:mo>-</mml:mo> <mml:mn>0.10</mml:mn> </mml:mrow> </mml:math> higher conventional regression models environments $$0.06-0.24$$ <mml:mn>0.06</mml:mn> <mml:mn>0.24</mml:mn> therefore datasets utilises becoming increasingly important emergence rapidly changing climate change.