Xiaoming Xie，Peng Zhao，Yuqi Zhang，Wenxi Wang，Zihao Wang， Zhaoxing Yu，Zhe Chen，Baoyue Zhang，Mingming Xin，Zhongfu Ni， Qixin Sun，Weilong Guo
DOI: 10.1016/j.xplc.2026.101841
Plant Communications
Abstract
As a cornerstone of global food security, wheat (Triticum aestivum L.) faces unprecedented pressure from a growing population and a changing climate, yet traditional breeding methods are increasingly insufficient to navigate the genetic complexity required for essential gains in yield and resilience. This review highlights key advances in generating large-scale, standardized datasets through the integration of high-throughput genotyping and multidimensional phenotyping. We explore how multi-omics integration and knowledge graph-based frameworks are crucial for converting this heterogeneous data into actionable breeding knowledge. We then examine the pivotal role of artificial intelligence (AI) and machine learning in driving predictive modeling, refining genomic selection, and enabling intelligent decision-making. These advances culminate in the emerging paradigm of Breeding 5.0, which harnesses data-driven innovation and closed-loop iterative cycles. Looking forward, we outline how multimodal AI and personalized breeding strategies will be critical for creating sustainable systems capable of ensuring global food security under climate change.
