Integrating genomic prediction into grapevine breeding programs

Abstract

Genomic selection (GS) has emerged as a transformative tool for accelerating breeding programs by predicting the genetic potential of individuals using genome-wide markers. In grapevine breeding, GS holds significant promise for enhancing genetic gains, reducing breeding cycles, and efficiently addressing complex traits such as yield, wine quality and climate adaptability. Despite its widespread success in animal breeding, the routine implementation of GS in plant breeding, remains limited due to challenges in prediction accuracy influenced by various factors.

We will share the first results from ongoing genomic prediction (GP) studies within the French breeding program INRAE-ResDur. This program aims to develop durable disease-resistant grapevine cultivars by integrating multiple sources of resistance. We conducted GP across several breeding populations, focusing on key phenological traits (e.g., flowering time, ripening date) and agronomic traits (e.g., berry weight, yield components, quality parameters). Our findings demonstrate that GP models can achieve moderate to high prediction accuracies for these traits, varying with the genetic architecture and design of the training population. We will also discuss how factors such as marker density, statistical models, and the inclusion of pedigree information influence the predictive ability of GP models.

To complement GP, we performed genome-wide association studies (GWAS) to identify significant loci and candidate genes controlling traits of interest. By incorporating these loci as fixed effects or employing marker-assisted selection alongside GS, we aim to enhance prediction accuracies and facilitate the introgression of desirable alleles. Integrating GS and GWAS into traditional breeding efforts offers a transformative approach to grapevine breeding. This combined strategy enables the evaluation of a larger number of seedlings at early stages, increasing selection efficiency and reducing the time and resources required for extensive field evaluations. Ultimately, applying genomics-based tools has the potential to accelerate the development of innovative cultivars better adapted to future challenges, such as emerging diseases and climate change. We will discuss practical considerations, potential limitations, and future directions for integrating these genomic tools into breeding strategies.