Jiayi Wu , Huijuan Xu , Qian Li

Modern Agricultural Research Center, Cuixi Academy of Biotechnology, Zhuji, 311800, Zhejiang, China
Author    Correspondence author
Maize Genomics and Genetics, 2025, Vol. 16, No. 3   
Received: 13 Apr., 2025    Accepted: 24 May, 2025    Published: 16 Jun., 2025

This is an open access article published under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Drought is one of the most severe abiotic stresses faced by maize (Zea mays L.) production worldwide, which seriously restricts the stability of crop yield. Traditional breeding methods have limited adaptability in the context of complex climate change, and more efficient prediction methods are urgently needed. This study integrates genomic selection (GS) and machine learning (ML) methods, and uses large-scale genotype, phenotype and environmental data to improve the accuracy of maize yield prediction under drought conditions. This article systematically reviews the latest progress in genomic prediction of drought resistance traits, analyzes typical machine learning algorithms suitable for crop modeling, and proposes a strategy for integrating GS and ML and a hybrid model framework construction method. The feasibility and practicality of this method are verified through actual cases such as the CIMMYT drought-resistant maize project and Chinese maize hybrids. At the same time, the model's portability and robustness in different ecological environments are also evaluated. This study provides a theoretical basis and technical path for AI-driven precision breeding, which has important guiding significance for the cultivation of new maize stress-resistant varieties under drought conditions.

Maize; Genomic selection; Machine learning; Drought stress; Yield