Comparative study of multi-trait genomic and phenotypic selection indexes for selection of superior genotypes

Genomic selection (GS) emerged as an efficient and cost-effective breeding technique that selects individuals based on their genetic merit via the prediction of genomic estimated breeding values (GEBVs) using molecular markers distributed over the entire genome. Genomic selection index (GSI) is a linear combination of GEBVs, while the phenotypic selection index (PSI) is a linear combination of multiple observable phenotypic traits. In this study, we compared the predictive performance of five parametric GS models such as RR-BLUP, Bayesian LASSO, Bayes A, Bayes B, and Bayes C for estimating GSI. Further, the GSI and PSI efficiency of breeding candidates was evaluated by applying suitable evaluation measures such as correlations of each indices with the net genetic merit, selection response, and expected genetic gain per trait. The findings of this study were further validated by two real datasets, suggesting that the GSI was more efficient than the PSI per unit of time.