Optimizing biomass production and antioxidant dynamics in Ceratophyllum demersum L. using multi-walled carbon nanotubes and machine learning models

In vitro plant culture systems are widely used to investigate biomass production with physiological regulation, and can be benefitted by incorporating novel materials such as nanoparticles. Aquatic macrophytes are highly significant for sustainable biomass production, with optimization remains a major challenge. In this study, the combined impact of multi-walled carbon nanotubes (MWCNTs), sucrose (S), and Murashige and Skoog (MS) basal salt concentrations on plant biomass, pigmentation, and biochemical activities in Ceratophyllum demersum was evaluated. The experiment was designed using Box-Behnken Design (BBD) of response surface methodology (RSM). The optimization results were further predicted and validated using machine learning models, specifically the Multilayer Perceptron (MLP) and Random Forest (RF) models, with six performance metrics, and the leave-one-out cross-validation technique. RSM analysis revealed a significant relationship between input parameters (MS basal salts, sucrose, and MWCNTs) and plant responses (growth and physiological traits). Among the testing ML models, the MLP model demonstrated superior predictive accuracy for several traits and varied for biochemical traits. Results illustrated the robust predictive capacity with R2 values of 0.99 for oxidative stress markers such as malondialdehyde and hydrogen peroxide. The integrated RSM-ML framework demonstrated robust predictive capacity of exidative markers and revealed distinct optimal conditions for growth and biochemical traits. The results provide valuable insights for studies conducted under in vitro conditions on how nanomaterials interact with nutritional and carbon sources for optimizing biomass production, physiological responses, and stress regulation in aquatic plants.