关键词: 作物, 全基因组选择, 全基因组预测模型, 育种

Abstract:

Genome selection（GS）breeding builds a genetic model based on the association between genotypes of molecular markers on the whole genome and phenotypes of the training population, and then estimates the breeding values or predicts the phenotypes of the candidate population with known genotypes, so as to achieve efficient and accurate selection of the population for breeding. Compared with the commonly used molecular marker-assisted selection breeding, GS breeding does not require marker significance testing, and is particularly suitable for quantitative traits controlled by minor polygenes. It can shorten breeding cycle and reduce breeding cost, and has become a cutting-edge technology in the field of animal and plant breeding. However, for quantitative traits such as crop yield that are greatly affected by environment, it is still bottleneck issue to improve the accuracy of genomic prediction. This article first analyzes the main factors that affect the efficacy of GS in crop breeding, and then elaborates on the research progress of GS technology in crop breeding from the aspects of models with non-additive effects, population construction schemes, multi-trait and multi-environment prediction, multi-omic prediction and the current status of breeding chip technology. Then the article points out the issues and development prospects of the research, and provides the strategies and ideas for further research on crop GS breeding technology.

Key words: crop, genomic selection, genomic prediction model, breeding