Determination of minimum number of random SNP for accurate population classification in rice (Oryza sativa L.) [post]

Shan Wang, Junhua Ye, Qun Xu, Xin Xu, Yingying Yang, Mengchen Zhang, Yue Feng, Xiaoping Yuan, Hanyong Yu, Yiping Wang, Xinghua Wei, Yaolong Yang
2020 unpublished
Background: Classification of germplasm collections is of great importance for both the conservation and utilization of genetic resources. Thus, it is necessary to estimate and classify rice varieties in order to utilize these germplasms more efficiently for rice breeding. However, molecular classification of large germplasm collections can be costly and labor-intensive. Development of an informative panel of a few markers would allow for rapid and cost-effective assignment of crops to genetic
more » ... f crops to genetic sub-populations.Results: Here, the minimum number of random SNP for rice classification (MNRSRC) was studied using a panel of 51 rice varieties belonging to different sub-groups. Through the genetic structure analysis, the rice panel can be obviously divided into five subgroups. The estimation of the MNRSRC was performed using SNP random sampling method based on genetic diversity and population structure analysis. In the genetic diversity analysis, statistical analysis of the coefficient of variation (CV) was performed for MNRSRC estimation, and we found that CV variation tended to plateau when the number of SNP was around 200, which was verified by the both cross-validation error of K value and correlation analysis of genetic distance. When the number of SNPs was greater than 200, the distribution of cross-validation error value tended to be similar, and correlation coefficients, almost greater than 0.95, exhibited small range of variation. In addition, we found that MNRSRC might not be affected by the number of varieties and the type of varieties.Conclusion: The estimation of the MNRSRC was performed using SNP random sampling method based on genetic diversity and population structure analysis. The results demonstrated that at least about 200 random filtered SNP loci were required for classification in a rice panel. In addition, we also found that MNRSRC might not be affected by the number of varieties and the type of varieties. The study on MNRSRC in this study can provide a reference and theoretical basis for classification of different types of rice panels.
doi:10.21203/rs.3.rs-35906/v1 fatcat:i64lgd5uxzbh3d5bocfc6b6zgy