Time-Resolved Chemical Phenotyping of Whole Plant Roots with Printed Electrochemical Sensors and Machine Learning

Abstract

Plants are non-equilibrium systems consisting of time-dependent biological processes. Phenotyping of chemical responses, however, is typically performed using plant tissues, which behave differently to whole plants, in one-off measurements. Single point measurements cannot capture the information rich time-resolved changes in chemical signals in plants associated with nutrient uptake, immunity or growth. In this work, we report a high-throughput, modular, real-time chemical phenotyping platform for continuous monitoring of chemical signals in the often-neglected root environment of whole plants: TETRIS ( T ime-resolved E lectrochemical T echnology for plant R oot I n-situ chemical S ensing). TETRIS consists of screen-printed electrochemical sensors for monitoring concentrations of salt, pH and H 2 O 2 in the root environment of whole plants. TETRIS can detect time-sensitive chemical signals and be operated in parallel through multiplexing to elucidate the overall chemical behavior of living plants. Using TETRIS, we determined the rates of uptake of a range of ions (including nutrients and heavy metals) in Brassica oleracea acephala. We also modulated ion uptake using the ion channel blocker LaCl 3 , which we could monitor using TETRIS. We developed a machine learning model to predict the rates of uptake of salts, both harmful and beneficial, demonstrating that TETRIS can be used for rapid mapping of ion uptake for new plant varieties. TETRIS has the potential to overcome the urgent “bottleneck” in high-throughput screening in producing high yielding plant varieties with improved resistance against stress.

Plant phenotyping relevance

植物全体の根圏における化学シグナルを連続測定する高スループット表現型解析プラットフォームを開発し、イオン吸収速度の推定と機械学習による予測まで行っており、フェノタイピング手法が研究の中心である。

Code and data availability

The supplied blocks describe the TETRIS electrochemical phenotyping platform, the salt-uptake dataset (~115 experiments), and the XGBoost machine learning model, but contain no data or code availability statement, no public repository, and no author-provided URL for datasets, sensor data, or trained models. The only 'n

No evidence-backed public reproduction asset is currently recorded.