Deep learning models for predicting plant uptake of emerging contaminants by including the role of plant macromolecular compositions

Abstract

Deep learning models can predict uptake of emerging contaminants in plants with improved accuracy because they leverage advanced data-driven approaches to capture non-linear relationships that traditional models struggle to address. Traditional models suffer from low accuracy in predicting transpiration stream concentration factor (TSCF) and root concentration factor (RCF). This study applied deep neural networks (DNN), recurrent neural networks (RNN), and long short-term memory (LSTM) to enhance the accuracy of predictive models for TSCF and RCF. The three models used nine chemical properties and two plant root macromolecular compositions for predicting TSCF and RCF. The results indicated that deep learning models predict TSCF and RCF with improved accuracy compared to mechanistic models. The coefficient of determination (R²) for the DNN, RNN, and LSTM models in predicting TSCF was 0.62, 0.67, and 0.56, respectively. The corresponding mean squared error (MSE) on the test set for the models was 0.055, 0.035, and 0.060, respectively. The R² for the DNN, RNN, and LSTM models in predicting RCF was 0.90, 0.91, and 0.84, respectively. The corresponding MSE for the models was 0.124, 0.071, and 0.126, respectively. The results of feature extraction using extreme gradient boosting underlined the importance of lipophilicity and root lipid fraction.

Keywords: Deep learning; Emerging contaminants; Machine learning; RCF; TSCF.