Designing Microbial Communities to Strengthen Plants

September 9, 2026

Machine learning enables the design and validation of microbial communities that promote tomato growth

Although many beneficial microorganisms that promote plant growth have been identified, it has been difficult to efficiently predict and design which combinations of microorganisms will produce the greatest effects.

A research group from Tohoku University, the RIKEN CSRS, and Kyoto University has developed a method for predicting and designing microbial communities that promote plant growth by integrating information on microbial combinations, microbial genetic information, plant-derived compounds, cultivation temperature, and other factors using machine learning. The designed microbial communities enhanced tomato growth under high-temperature conditions in the laboratory and in outdoor agricultural field conditions. Furthermore, through multi-omics analyses, the researchers identified plant genes and microorganisms associated with these effects, shedding light on part of the underlying mechanisms.

This achievement demonstrates a new approach that goes beyond simply “finding” beneficial microorganisms to “designing” effective microbial communities based on data. The approach is expected to contribute to the development of agricultural systems that are more resilient to environmental fluctuations.

 

Original article

The ISME Journal　DOI: 10.1093/ismejo/wrag219

S. Yamazaki, M. Nakayasu, K. Kanai, R. Mizuno, R. Kaida, S. Masuda, A. Shibata, K. Shirasu, A. J. Nagano, Y. Fujii, A. Sugiyama, Y. Aoki,

"Model-guided design of defined microbial community reveals interactions underpinning plant growth and stress tolerance".

Contact

Shinichi Yamazaki; Research Scientist
Holobiont and Resilience Research Team

Ken Shirasu; Group Director
Sachiko Masuda; Research Scientist
Plant Immunity Research Group