Application of multi-trait Bayesian decision theory for parental genomic selection

Abstract

In all breeding programs, the decision about which individuals to select and intermate to form the next selection cycle is crucial. The improvement of genetic stocks requires considering multiple traits simultaneously, given that economic value and net genetic merits depend on many traits; therefore, with the advance of computational and statistical tools and genomic selection (GS), researchers are focusing on multi-trait selection. Selection of the best individuals is difficult, especially in traits that are antagonistically correlated, where improvement in one trait might imply a reduction in other(s). There are approaches that facilitate multi-trait selection, and recently a Bayesian decision theory (BDT) has been proposed. Parental selection using BDT has the potential to be effective in multi-trait selection given that it summarizes all relevant quantitative genetic concepts such as heritability, response to selection and the structure of dependence between traits (correlation). In this study, we applied BDT to provide a treatment for the complexity of multi-trait parental selection using three multivariate loss functions (LF), Kullback-Leibler (KL), Energy Score, and Multivariate Asymmetric Loss (MALF), to select the best performing parents for the next breeding cycle in two extensive real wheat data sets. Results show that the high ranking lines in genomic estimated breeding value (GEBV) for certain traits did not always have low values for PEL. For both data sets, the KL loss function gave similar importance to all traits including grain yield. In contrast, the Energy Score and MALF gave better performance in 3 of 4 traits that were different than grain yield. The BDT approach should help breeders to decide based not only on the GEBV per se of the parent to be selected, but also on the level of uncertainty according to the Bayesian paradigm.

A Bayesian optimization R package for multitrait parental selection

“…The theory of LFs, used to enhance genetic gain in selection programs, was developed in our previous publications (Villar‐Hernández et al., 2018, 2021). The foundation of LF theory stems from the concept of selection by truncation in single traits and extends to address multitrait selection.…”

“…Following the initial investigation, more recently Villar‐Hernández et al. (2021) conducted multitrait selection using real wheat datasets with four traits and employed the BDT framework and the concept of LFs to explore practical applications and address gaps in a previous study (Villar‐Hernández et al., 2018). The three LFs used for multitrait selection were KL, EnergyS, and MALF.…”

“…The theory of our approach can be found in previous publications of Villar‐Hernández et al. (2018, 2021). Our goal is to develop a package that can be integrated into the Bayesian Generalized Linear Regression (BGLR) software (Pérez‐Rodríguez & de los Campos, 2022), which fits many genomic prediction models and enjoys broad acceptance within the breeding community.…”

Genomic prediction based on a joint reference population for the Xinjiang Brown cattle

“…In theory, a model that assumes the closest distribution of SNP effects to their true distribution can achieve the highest reliability in genomic prediction. The GBLUP model assumes that all SNP effects follow the same normal distribution and compresses the effects of all SNPs to the same degree because different models have different assumptions about the distribution of SNP effects (Villar-hernÁndez et al, 2021). Several methods have been proposed to improve the accuracy of genomic prediction in small populations of dairy cattle (Marjanovic et al, 2021), and one effective approach is to use joint reference populations by combining reference data from different populations (Steyn et al, 2019;van Grevenhof et al, 2019).…”

Multi‐trait genomic prediction in pigs using single and multistep methods based on the absorption of ungenotyped animals

“…Actual breeding often targets multiple traits that are genetically correlated, and the practical routine genetic evaluation of the breeding value is usually calculated using multi‐trait models. Multi‐trait models for GEBV prediction have been reported including Bayesian approaches (Villar‐Hernandez et al, 2021) and the genomic best linear unbiased prediction (GBLUP) method (Karaman et al, 2020). Studies have shown that multi‐trait genomic prediction (MTGP), which accounts for the relationships between the traits, may result in more accurate GEBV than single‐trait genomic prediction (STGP) (Semagn et al, 2022; Song et al, 2020).…”