Population genomics identifies the origin and signatures of selection of Korean weedy rice

Abstract

Weedy rice is the same biological species as cultivated rice (Oryza sativa); it is also a noxious weed infesting rice fields worldwide. Its formation and population-selective or -adaptive signatures are poorly understood. In this study, we investigated the phylogenetics, population structure and signatures of selection of Korean weedy rice by determining the whole genomes of 30 weedy rice, 30 landrace rice and ten wild rice samples. The phylogenetic tree and results of ancestry inference study clearly showed that the genetic distance of Korean weedy rice was far from the wild rice and near with cultivated rice. Furthermore, 537 genes showed evidence of recent positive or divergent selection, consistent with some adaptive traits. This study indicates that Korean weedy rice originated from hybridization of modern indica/indica or japonica/japonica rather than wild rice. Moreover, weedy rice is not only a notorious weed in rice fields, but also contains many untapped valuable traits or haplotypes that may be a useful genetic resource for improving cultivated rice.

Keywords: genetic resources; natural selection; origin; weedy rice; whole-genome resequencing.

Figure 1

Population structure analysis of 30 weedy rice samples with 40 other Oryza species. (a) Maximum likelihood phylogenetic tree. Blue branches are weedy rice, black branches are landrace rice. (b) Maximum likelihood clustering with K ranging from 2 to 7. For each K, the different colour represents different populations. Each accession is represented by a vertical bar, and the length of each coloured segment in each vertical bar represents the proportion contributed by ancestral populations. (c) PCA plots. Blue dots are weedy rice, black dots are landrace rice. (d) Illustration of genetic diversity and population differentiation in wild rice and japonica_landrace, japonica_weedy, indica_landrace and indica_weedy. The sizes of the circles represent the levels of genetic diversity of groups, and the distances are F _ST values between different groups.

Figure 2

Unique and shared genomic regions (genes) among five selection scans. (a) Venn diagram of the number of genes throughout the genome in the top 2.5% for each selection scan. (b) GO enrichment of candidate selection genes. The blue bar indicates candidate selection genes, and the green bar shows the rice reference genome genes (c) KEGG pathway enrichment of candidate selection genes.

Figure 3

Candidate selected genes clustered based on public databases. (a) Functional category of candidate selected genes. (b) Venn diagram of three functional category. (c) The ‘Biotic Stress overiew’ MapMan pathway was used to visualize candidate genes. Red square are genes. (d) Genes involved in receptor‐like kinases. Red square are genes.