Résumé
This chapter focuses on analysing factors that affect the predictive ability of genomic prediction (GS) in rice, and an overview of the proof-of-concept studies conducted during the past 5 years with the aim of providing rice breeders with tailored GS methods and tools. These studies involved two complementary breeding schemes (pedigree breeding and population improvement), mobilized different compartments of the rice genetic diversity (indica, tropical japonica and temperate japonica) and targeted a wide range of traits (yield potential, adaptation to alternate watering and drying, drought tolerance and exclusion of heavy metals). Issues addressed include training the population for making selection decisions in pedigree breeding within the progeny of biparental crosses, accounting for information available on gene/QTL involved in the determination of complex traits and accounting for genotype-by-environment interactions. In the light of the results of these studies, a strategy for the implementation of genomic selection in the framework of pedigree breeding was discussed. Issues that need further simulation and empirical studies to fully harness the potential of GS was also discussed.