人工智能驱动的植物科学跨尺度融合研究:进展、挑战与展望OA

The rapid development of multi-omics technologies,high-throughput phenotyping technologies,and agricul-tural and ecosystem observation networks has generated massive,heterogeneous,and multi-scale datasets,spanning multiple layers of plant biological systems.Integrating these heterogeneous data to decipher complex biological proc-esses,and enable accurate prediction has become a major challenge in modern plant science.In recent years,artificial intelligence(AI)has emerged as a powerful approach for multi-omics integration,gene function prediction,plant pheno-typing,and modeling of complex biological processes,providing powerful tools and new perspectives for plant research.This review systematically summarizes recent advances in AI applications across multiple organizational levels of plant systems,covering molecular regulation,trait formation,environmental adaptation,agricultural production,and ecosystem research.Emphasis is placed on the roles of AI in cross-scale data integration,deciphering complex biological processes,and predictive modeling.We further discuss current challenges,including data quality and sharing,model interpretability,cross-scale coupling,and insufficient mechanistic understanding.Looking forward,AI frameworks that integrate data-driven approaches with biological knowledge and mechanistic models are expected to advance plant science from descriptive and correlative analyses toward mechanistic understanding,functional prediction,and intelligent design,thereby providing a stronger theoretical and technological foundation for basic plant research,smart agriculture,and the sustainable utilization of plant resources.