Integrating genome-wide association study into genomic selection for the prediction of agronomic traits in rice (Oryza sativa L.)

• DOI: 10.1007/s11032-023-01423-y
• Publication Date: 2023-11-13T02:02:02Z
• Authors: Yuanyuan Zhang; Mengchen Zhang; Junhua Ye; Qun Xu; Yue Feng; Siliang Xu; Dongxiu Hu; Xinghua Wei; Peisong Hu; Yaolong Yang
• Keywords: Genome-wide Association Study; Genomic Selection; Genetic Association; Predictive modelling

Accurately identifying varieties with targeted agronomic traits was thought to contribute genetic selection and accelerate rice breeding progress. Genomic (GS) is a promising technique that uses markers covering the whole genome predict genomic-estimated values (GEBV), ability select before phenotypes are measured. To choose appropriate GS models for work, we analyzed predictability of nine measured from population 459 diverse varieties. By comparison eight representative models, found prediction accuracies ranged 0.407 0.896, reproducing kernel Hilbert space (RKHS) having highest predictive in most traits. Further results demonstrated predictivity altered by several factors. Moreover, assessed method integrating genome-wide association study (GWAS) into various models. The predictabilities combined peak-associated generated six different GWAS were significantly different; recommendation Mixed Linear Model (MLM)-RKHS given GWAS-GS-integrated prediction. Finally, based on above result, experimented applying P-values obtained optimal ridge regression best linear unbiased (rrBLUP), which benefited low rice.The online version contains supplementary material available at 10.1007/s11032-023-01423-y.