Artificial Intelligence and Machine Learning for Genomic Prediction, High-Throughput Phenotyping and Climate-Adaptive Breeding In Maize and Rice: A Comprehensive Review

Abstract

Climate change is intensifying abiotic stresses such as drought and heat, posing significant threats to global food security and the productivity of staple crops including maize (Zea mays L.) and rice (Oryza sativa L.). Conventional breeding approaches are often constrained by the complex genetic architecture of stress-adaptive traits and lengthy breeding cycles, highlighting the need for more efficient, data-driven strategies. This review summarizes recent advances in artificial intelligence (AI) and machine learning (ML) for genomic prediction, high-throughput phenotyping (HTP), and climate-adaptive breeding in maize and rice. We discuss the applications of machine learning architectures, including multilayer perceptron (MLP), convolutional neural networks (CNN), random forest (RF), deep neural networks (DNN), gradient boosting methods, and explainable artificial intelligence (XAI), in improving genomic selection and capturing complex genotype–environment interactions. The review further explores the integration of AI with HTP technologies, including autonomous robotic platforms, drones, hyperspectral imaging, and LiDAR, to enable rapid, accurate, and non-destructive phenotypic assessment. In addition, we examine the role of AI-driven predictive models in identifying stress-responsive genes, improving trait prediction, and accelerating the development of climate-resilient crop varieties. Current challenges, including data heterogeneity, computational demands, model interpretability, and biological validation, are also discussed alongside emerging solutions such as multi-view learning, transfer learning, and intelligent precision design breeding. Overall, the convergence of AI, ML, multi-omics, and advanced phenotyping technologies represents a transformative framework for next-generation crop improvement, offering new opportunities to accelerate sustainable breeding programs and strengthen global food security under changing climatic conditions.

Plant phenotyping relevance

AI・MLを用いた高スループット植物表現型解析と、ロボット、ドローン、ハイパースペクトル、LiDARによる表現型評価を中心的にレビューしているため。

Code and data availability

公開論文であることは確認できましたが、現在の公式API・許可済み取得経路では本文を自動取得できませんでした。

No evidence-backed public reproduction asset is currently recorded.