A QSTR model for toxicity prediction of pesticides towards Daphnia magna

Because of the large amount of pesticides discharged into rivers, adverse effects could be induced to aquatic organisms. Daphnia magna is often used as an indicator organism to evaluate the toxicity of pesticides. In this study, a quantitative structure-toxicity relationship (QSTR) model was established based on norm descriptors for predicting the acute toxicity of pesticides to Daphnia magna. The model results showed the good predictability (R-training(2) = 0.8045, R-testing(2) = 0.8224). The validation results of internal validation, external validation, Y randomization test and application domain analysis demonstrated the model's stability, reliability and robustness. Therefore, the above results indicate that norm descriptors might be universal for describing the relationship between the toxicity and structures of pesticides compounds. Moreover, some pesticides' toxicities without experimental data were also predicted by this model.

Automated summary

A quantitative structure-toxicity relationship (QSTR) model based on norm descriptors was established in this study to predict the toxicity of pesticides to Daphnia magna. The model showed good predictability, stability, and reliability, indicating that norm descriptors might be universally applicable in describing the relationship between the toxicity and structures of pesticide compounds.