Rich Genomic and Agro-Morphological Diversity Provides Fuel for Rice Varietal Improvement.

Cultivated Asian rice (Oryza sativa L.) has tremendous phenotypic and genotypic diversity that traces back to ancient times. A “Rice Diversity Panel” composed of 413 diverse rice accessions from 82 countries was assembled to explore this genotypic and phenotypic variation. The objectives of this study were to purify the accessions in the panel, collect phenotypic data from field-grown plants in a replicated study; genotype the accessions with 44,100 (44K) single nucleotide polymorphism (SNP) markers and later with 700K SNPs; conduct genome-wide association studies (GWAS); and enter the phenotypic and genotypic data into appropriate databases.  Population structure based on 44K SNPs using principle component analysis identified the five rice subpopulations (indica, aus, aromatic, tropical japonica, temperate japonica); some accessions were admixtures of two or more subpopulation groups. Canonical discriminant analysis of the phenotypic data confirmed a similar population structure. Results of GWAS identified different marker-trait associations for the same traits in each of the subpopulations, suggesting that different combinations of alleles were responsible for the phenotypic variation observed in each subpopulation. This helps explain the transgressive variation that is observed in crosses between subpopulations and has immediate implications for rice breeding. The most significant genotype-phenotype associations were observed for days to heading, plant height, panicle morphology, grain shape, pericarp color, amylose content, starch gelatinization temperature, and rice blast disease. A UCSC browser was developed to integrate gene and genome annotations with the SNP data generated on this rice diversity panel. The project website (ricediversity.org) hosts the genotypic and phenotypic data, and a “Seed Photo Library” of all accessions; GRIN hosts panicle pictures, SSR data, and most phenotypic data.