Prediction of Plant Uptake and Translocation of Engineered Metallic Nanoparticles by Machine Learning

• DOI: 10.1021/acs.est.1c01603
• Publication Date: 2021-05-17T19:13:41Z
• Authors: Xiaoxuan Wang; Liwei Liu; Weilan Zhang; Xingmao Ma
• Keywords: Plant Roots; Natural Organic Matter

Machine learning was applied to predict the plant uptake and transport of engineered nanoparticles (ENPs). A back propagation neural network (BPNN) used root concentration factor (RCF) translocation (TF) ENPs from their essential physicochemical properties (e.g., composition size) key external factors exposure time species). The relative importance input variables determined by sensitivity analysis, gene-expression programming (GEP) generate predictive equations. BPNN model satisfactorily predicted RCF TF in both hydroponic soil systems, with an R2 higher than 0.8 for all simulations. Inclusion initial ENP as variable further improved accuracy systems. Sensitivity analysis indicated that metals vs metal oxides) is a major affecting values system. However, organic matter clay contents are more dominant GEP (R2 = 0.8088 0.8959 values) generated accurate equations conventional regression 0.5549 0.6664 system, which could guide sustainable design agricultural applications.