WITHDRAWN: Differentially directional selection on grain size genes identified by RapMap during rice domestication

Abstract The ability to identify quantitative trait locus (QTL) genes controlling natural variation of important traits would facilitate the construction of a bridge between gene function analyses and domestication investigations. Grain size in rice is an important yield, quality and domestication trait. The molecular architecture and domestication signatures of grain size at population level remain unknown. Here we first introduce RapMap, a method for rapid and high-throughput QTL mapping employing a series of F 2 gradient populations (F 2 GPs) constructed by trait-adjacent accessions from diverse germplasms. A co-segregation standard is set for simultaneously integrating the three-in-one framework in RapMap: detecting a real QTL, confirming its effect and obtaining its near-isogenic line-like line (NIL-LL), which greatly enhances the efficiency and robustness of QTL identification. Using 15 gradient populations, eight genes including two novel genes (GL1 and GW5.1 ) with different effects for grain size in rice were simultaneously discovered by RapMap in three years, which make more than two thirds of genetic contributions to grain size and shape in a mini-core collection worldwide. In-depth analyses of the eight genes employing a large and geographically diverse population, including 446 wild, 2462 landrace and 784 cultivar accessions, revealed that directional artificial selection of slender- and long- grains and their alleles played an important role during indica rice domestication. The genetic effects of the eight grain-size genes positively correlate with the differential selection intensities and the nucleotide variation intensities. Major grain-size genes fixed before the completion of rice domestication have been heavily selected for human requirements and cultivation practices. Generalizable application of the method and mimicking of the domestication nature will accelerate breeding modern cultivars.