Clustering strategy for establishing population structure of 3024 rice genomes based on single nucleotide polymorphism
2017
Vasquez, M.S.C.
Inference of population structure and classification of different rice accessions based on genetic and phenotypic information are essential in the study of the natural history of the organism and of the domestication practices of humans. In this study, a model-based clustering method was used to cluster 3,024 rice accessions based on their Single Nucleotide Polymorphisms (SNPs). The procedure partitioned the accessions into 12 clusters, namely aromatic, aus, indica 1, indica 2, indica 3, indica 4, indica 5, subtropical, temperate, tropical A-1, tropical A-2, and tropical B. Many of these groups were observed to be associated to specific geographic locations. It was also found that observations classified as temperate have characteristics that are distinct from the other clusters. The hierarchical clustering approach using Unweighted Pair Group Method with Arithmetic Mean (UPGMA) was performed to cluster the accessions according to their phenotypes. Four cluster validation indexes suggested that the rice accessions can be divided into two or four subpopulations. Through Random Forest, this paper was able to illustrate that varietal groups formed based on the SNPs can be predicted using the phenotypes. Important variables for classification include flowering time, culm length, and grain length.
显示更多 [+] 显示较少 [-]