Implementing marker covariates and multi-trait genomic selection models to improve grain milling, appearance, cooking, and edible quality in rice (Oryza sativa L.).

Abstract

Rice (Oryza sativa L.) is a staple food for over half of the world's population. With population growth, socioeconomic changes, and shifting consumer lifestyles, the demand for high-quality rice has surged. Understanding consumer preferences for rice quality traits is crucial for breeders to effectively address evolving market needs. Rice breeding programs assess various quality aspects, including grain shape, appearance, milling efficiency, and cooking and eating qualities. Molecular-based approaches like marker-assisted selection and genomic selection (GS) offer promising opportunities to enhance breeding efficiency. In this study, our goal was to build upon our previous findings and improve the predictive ability of GS for primary grain milling and cooking and eating quality traits by incorporating trait marker covariates and highly heritable, high-throughput secondary traits in multi-trait genomic selection strategies (MT-GS). By including amylose content and gelatinization temperature functional markers as covariates in GS models, we improved the predictive ability for primary cooking and eating traits from 21% to 44%. Additionally, integrating secondary traits into MT-GS increased the predictive ability for milling quality traits from 13.5% to 18% and for cooking and eating traits from 4.6% to 50%. Overall, our study demonstrates the feasibility of incorporating whole-genome markers, trait markers, and secondary trait information to enhance the predictive ability of GS for grain milling, cooking, and eating qualities in rice.