基于一般配合力的杂交种全基因组选择方法研究

In the field of crop breeding, accurately predicting the phenotypes of the F1 generation of hybrids is crucial for cultivating superior varieties and enhancing breeding efficiency. How to break through the existing predicament and improve the prediction accuracy of low-heritability traits to provide more powerful technical support for the prediction of crop hybrids is currently the focus of the research and development of genome-wide selection methods. This study took two sets of hybrid populations, including a rice population with 925 hybrids and a maize population with 6210 hybrids, as the research objects. By using eight prediction methods, the general combining ability (GCA), genetic distance (GD), and parental phenotypic values (P) were creatively added as fixed effects to the prediction model to evaluate the feasibility of incorporating parental GCA into the hybrid prediction model. The results showed that compared with the SNP model, the SNP+GCA model could significantly improve the prediction accuracy of hybrid traits; different methods had different prediction accuracies for different hybrid traits, but all methods had significant improvements in prediction accuracy and stability after incorporating GCA information. The development of this new method has improved the prediction efficiency of hybrids, reduced time and resource costs, brought new ideas and tools for the selection and breeding of crop hybrids, and accelerated the breeding process of hybrids.