Proc. Natl Acad. Sci. USA http://doi.org/7st (2015)

Recently large-scale population genomic analyses have been effectively applied in the study of rice domestication. However, this powerful tool has yet to be sufficiently employed in the study of modern rice breeding. Weibo Xie and colleagues from the Huazhong Agricultural University have now unravelled the breeding signatures of indica rice by analysing the genomes of 1,479 rice accessions.

Using the genomic data from two previous studies, the researchers obtained over six million single nucleotide polymorphisms (SNPs) classifying indica rice into two groups — IndI and IndII. About 37% of the IndI accessions and 95% of the IndII accessions are semi-dwarf improved cultivars, corresponding to two heterotic groups. The founder lines of the two heterotic groups were inferred to be Aijiaonante and IR8, respectively.

Selective sweep analysis showed that 200 regions, covering 7.8% of the rice genome, have been differentially selected in the two groups, some of which overlap with domestication-related regions. Genes related to important agronomic traits — such as plant architecture, disease resistance and grain yield — are found in these regions.

A significant correlation was observed between yield and the number of selected SNP alleles or haplotypes in an accession in both the indica population and another two artificial segregating populations. These selected alleles or haplotypes are breeding signatures that can be used to predict agronomic potential and facilitate rice improvement.