Whole Genome Sequencing and Morphological Trait-Based Evaluation of UPOV Option 2 for DUS Testing in Rice

Hong Liu; Dehua Rao; Tao Guo; Sunil S Gangurde; Yanbin Hong; Mengqiang Chen; Zhanquan Huang; Yuan Jiang; Zhenjiang Xu; Zhiqiang Chen

doi:10.3389/fgene.2022.945015

Whole Genome Sequencing and Morphological Trait-Based Evaluation of UPOV Option 2 for DUS Testing in Rice

Front Genet. 2022 Aug 26:13:945015. doi: 10.3389/fgene.2022.945015. eCollection 2022.

Authors

Hong Liu^{1

2}, Dehua Rao², Tao Guo¹, Sunil S Gangurde^{3

4

5}, Yanbin Hong⁶, Mengqiang Chen², Zhanquan Huang², Yuan Jiang², Zhenjiang Xu², Zhiqiang Chen¹

Affiliations

¹ National Engineering Research Center of Plant Space Breeding, South China Agricultural University, Guangzhou, Guangdong, China.
² College of Agriculture, South China Agricultural University, Guangzhou, Guangdong, China.
³ International Crops Research Institute for the Semi-Arid Tropics (ICRISAT), Hyderabad, India.
⁴ Crop Protection and Management Research Unit, USDA-ARS, Tifton, GA, United States.
⁵ Department of Plant Pathology, University of Georgia, Tifton, GA, United States.
⁶ Guangdong Provincial Key Laboratory for Crops Genetic Improvement, Crops Research Institute, Guangdong Academy of Agricultural Sciences, Guangzhou, China.

Abstract

To evaluate the application potential of high-density SNPs in rice distinctness, uniformity, and stability (DUS) testing, we screened 37,929 SNP loci distributed on 12 rice chromosomes based on whole-genome resequencing of 122 rice accessions. These SNP loci were used to analyze the DUS testing of rice varieties based on the correlation between the molecular and phenotypic distances of varieties according to UPOV option 2. The results showed that statistical algorithms and the number of phenotypic traits and SNP loci all affected the correlation between the molecular and phenotypic distances of rice varieties. Relative to the other nine algorithms, the Jaccard similarity algorithm had the highest correlation of 0.6587. Both the number of SNPs and the number of phenotypes had a ceiling effect on the correlation between the molecular and phenotypic distances of varieties, and the ceiling effect of the number of SNP loci was more obvious. To overcome the correlation bottleneck, we used the genome-wide prediction method to predict 30 phenotypic traits and found that the prediction accuracy of some traits, such as the basal sheath anthocyanin color, glume length, and intensity of the green color of the leaf blade, was very low. In combination with group comparison analysis, we found that the key to overcoming the ceiling effect of correlation was to improve the resolution of traits with low predictive values. In addition, we also performed distinctness testing on rice varieties by using the molecular distance and phenotypic distance, and we found that there were large differences between the two methods, indicating that UPOV option 2 alone cannot replace the traditional phenotypic DUS testing. However, genotype and phenotype analysis together can increase the efficiency of DUS testing.

Keywords: DUS; SNP; correlation analysis; distinctness; genomic prediction; genotype; phenotype; rice.