A review of AI-driven phenomics, genomics, and predictive breeding in wheat
- 1Division of Plant Science and Technology, University of Missouri, Columbia, MO, USA
- 2Department of Plant and Soil Science, Texas Tech University, Lubbock, TX, USA
- 3Department of Plant Breeding and Genetics, Punjab Agricultural University, Ludhiana, Punjab, India
Abstract
Wheat ( Triticum aestivum L.) is a staple crop that feeds billions of people worldwide, but increasing pressure from climate change, pests, diseases, and growing food demands is making its improvement more urgent than ever. Traditional breeding methods, though valuable, often fall short in meeting these complex challenges quickly and efficiently. In recent years, artificial intelligence (AI), particularly machine learning (ML), has emerged as a powerful tool that is reshaping wheat breeding. While AI, and specifically ML approaches have advanced substantially in crop improvement, their application in wheat breeding remains fragmented, with limited integration across phenotyping, genotyping, and multi-omics layers, and insufficient attention to model interpretability and deployment in resource-constrained breeding programs. This review critically synthesizes recent ML-based approaches applied throughout modern wheat improvement, from large-scale phenotypic and genotypic data analysis to trait prediction and multi-omics integration. We highlight how these methods are being used to enhance genetic gain for resistance to biotic stress (e.g., disease and pests) and abiotic stress (e.g., heat and drought), while identifying key methodological and infrastructural barriers that currently limit their broader adoption. The role of big data, effective data management, and the difficulties in handling complex datasets are also discussed. Despite their promise, challenges remain, such as the need for better model transparency, access to computing resources in developing regions, and ensuring that these technologies are equitably accessible. Further, the review briefly discusses emerging directions, including deep learning (DL)-based frameworks, explainable AI (XAI), and their potential integration with genome editing, and digital twin concepts, as future opportunities to further enhance wheat breeding efficiency. Finally, we emphasize the importance of collaboration and open-access platforms to ensure these innovations benefit researchers and farmers alike. Together, these advancements mark a significant shift toward faster, smarter, and more sustainable wheat improvement for the future.
Related
- Omics-driven plant breeding through phenomics-enviromics crosstalk
- Integrating Envirotyping and Phenomics for AI-Enabled Multi-Environment Genomic Prediction in Crop Breeding
- From Gene-Based Breeding to Haplotype-Based Breeding for Rice (Oryza sativa) Improvement in the Genomics Era
- Agronomic characterisation of a common buckwheat germplasm collection unlocks resources for genomics-assisted ideotype breeding
- Pan Genome to Pan Phenome provides a mechanistic framework for translating hidden genetic diversity into predictive wheat breeding
BibTeX
@article{tanin2026review,
title = {{A review of AI-driven phenomics, genomics, and predictive breeding in wheat}},
author = {Mohammad Jafar Tanin and Dinesh Kumar Saini and Praveenkumar Alagappan and Jatinder Paul Kaur},
journal = {Discover Plants.},
year = {2026},
month = jun,
doi = {10.1007/s44372-026-00696-9},
url = {https://doi.org/10.1007/s44372-026-00696-9},
note = {CC BY-NC-ND},
}
RIS
TY - JOUR
TI - A review of AI-driven phenomics, genomics, and predictive breeding in wheat
AU - Mohammad Jafar Tanin
AU - Dinesh Kumar Saini
AU - Praveenkumar Alagappan
AU - Jatinder Paul Kaur
PY - 2026
DA - 2026/06/11
JO - Discover Plants.
DO - 10.1007/s44372-026-00696-9
UR - https://doi.org/10.1007/s44372-026-00696-9
AB - Wheat ( Triticum aestivum L.) is a staple crop that feeds billions of people worldwide, but increasing pressure from climate change, pests, diseases, and growing food demands is making its improvement more urgent than ever. Traditional breeding methods, though valuable, often fall short in meeting these complex challenges quickly and efficiently. In recent years, artificial intelligence (AI), particularly machine learning (ML), has emerged as a powerful tool that is reshaping wheat breeding. While AI, and specifically ML approaches have advanced substantially in crop improvement, their application in wheat breeding remains fragmented, with limited integration across phenotyping, genotyping, and multi-omics layers, and insufficient attention to model interpretability and deployment in resource-constrained breeding programs. This review critically synthesizes recent ML-based approaches applied throughout modern wheat improvement, from large-scale phenotypic and genotypic data analysis to trait prediction and multi-omics integration. We highlight how these methods are being used to enhance genetic gain for resistance to biotic stress (e.g., disease and pests) and abiotic stress (e.g., heat and drought), while identifying key methodological and infrastructural barriers that currently limit their broader adoption. The role of big data, effective data management, and the difficulties in handling complex datasets are also discussed. Despite their promise, challenges remain, such as the need for better model transparency, access to computing resources in developing regions, and ensuring that these technologies are equitably accessible. Further, the review briefly discusses emerging directions, including deep learning (DL)-based frameworks, explainable AI (XAI), and their potential integration with genome editing, and digital twin concepts, as future opportunities to further enhance wheat breeding efficiency. Finally, we emphasize the importance of collaboration and open-access platforms to ensure these innovations benefit researchers and farmers alike. Together, these advancements mark a significant shift toward faster, smarter, and more sustainable wheat improvement for the future.
SN - 3005-1207
ER -