生物技术通报 ›› 2026, Vol. 42 ›› Issue (4): 1-16.doi: 10.13560/j.cnki.biotech.bull.1985.2025-1157

• 综述与专论 •    下一篇

全球作物智能设计育种政策、技术及发展建议

林巧1(), 鲜国建1,2, 李慧慧3, 张学福1,2(), 孙坦2,4()   

  1. 1.中国农业科学院农业信息研究所,北京 100081
    2.农业农村部农业大数据重点实验室,北京 100081
    3.中国农业科学院作物科学 研究所,北京 100081
    4.中国农业科学院,北京 100081
  • 收稿日期:2025-10-27 出版日期:2026-04-26 发布日期:2026-04-30
  • 通讯作者: 张学福,男,博士,研究员,研究方向 :信息资源管理;E-mail: zhangxuefu@caas.cn
    孙坦,男,博士,研究馆员,研究方向 :数字信息描述与组织;E-mail: suntan@caas.cn
  • 作者简介:林巧,女,博士,副研究馆员,研究方向 :农业信息管理;E-mail: linqiao@caas.cn
  • 基金资助:
    国家社会科学基金项目(23BTQ054);中国农业科学院农业信息研究所基本科研项目(JBYW-AII-2024-20)

Global Strategies, Technologies, and Policy Recommendations for Crop Intelligent Design and Breeding

LIN Qiao1(), XIAN Guo-jian1,2, LI Hui-hui3, ZHANG Xue-fu1,2(), SUN Tan2,4()   

  1. 1.Agricultural Information Institute, Chinese Academy of Agricultural Sciences, Beijing 100081
    2.Key Laboratory of Agricultural Big Data, Ministry of Agriculture and Rural Affairs, Beijing 100081
    3.Institute of Crop Sciences, Chinese Academy of Agricultural Sciences, Beijing 100081
    4.Chinese Academy of Agricultural Sciences, Beijing 100081
  • Received:2025-10-27 Published:2026-04-26 Online:2026-04-30

摘要:

种子是农业不可替代的重要生产物质,智能设计育种通过运用大数据、人工智能和基因编辑等技术,构建与模拟作物从育种到收获的全过程,帮助实现作物育种从“试验优选”向“计算优选”的根本转变。本文概述了作物育种从驯化育种到智能设计育种的演进过程,基于科技论文,通过VOSViewer软件和LDA主题模型挖掘了全球智能设计育种领域的研究主题和技术方向,并深入探讨美国等农业发达国家在育种政策上的布局及其特点,总结中国在智能设计育种领域面临的问题和挑战,并提出相应的发展建议。研究表明:(1)全球智能设计育种相关研究共涉及三大主题九大技术方向,人工智能表型识别技术体系、智能计算育种技术体系、基因型-环境-表型适配性关联技术体系成为各国的关注重点。(2)国际上相关政策已明显体现出对育种4.0理念的重视,强调整合基因编辑、大数据和人工智能等高科技手段;我国目前的政策和研究重点仍主要集中在分子育种阶段,相关政策布局更侧重于种质资源保护、种源关键技术攻关、传统育种技术提升等方面。(3)综合我国智能设计育种在政策制定、核心技术、成果转化等方面面临的挑战,提出强化顶层设计、加快技术创新驱动、完善创新生态建设的发展建议。

关键词: 智能设计育种, 关键技术, 表型预测, 政策布局, 创新生态

Abstract:

Seed is a crucial, irreplaceable component in agriculture. Intelligent design and breeding employs technologies such as big data, artificial intelligence, and gene editing to model and simulate the entire crop growth process from breeding to harvest. This marks a substantial shift from traditional “experimental optimization” to “computational optimization”. This paper traces the evolution of crop breeding from its early domestication stages to the era of intelligent design and breeding. Using VOSViewer software and the LDA topic model, we analyze scientific literature to identify global research themes and technological trends inintelligent design and breeding. Additionally, we further discuss the layout and its characteristics of the breeding policies in agriculturally advanced countries, such as the United States. We also address the issues and challenges faced by China in this field and offer development recommendations. The study reveals that: 1) Global research in intelligent design and breeding focuses on three main themes and nine major technological directions, particularly AI phenotyping technology systems, intelligent computational breeding technology systems, and genotype-environment-phenotype fitness associating technology systems. 2) International policies emphasize the importance of the Breeding 4.0 concept, integrating high technologies such as gene editing, big data, and artificial intelligence. In contrast, China’s policies are predominantly centered on the molecular breeding stage, with a strong focus on germplasm resource protection, key technological breakthroughs in seed sources, and enhancements of traditional breeding methods. 3) To address the challenges in policy formulation, core technology development, and the application of research in intelligent design and breeding, we propose that China should strengthen top-level design, accelerate technology-driven innovation, and improve the innovation ecosystem.

Key words: intelligent design and breeding, key technologies, phenotypic prediction, policy frameworks, innovation ecosystem