Linear regression model for predicting interactive mixture toxicity of pesticide and ionic liquid

摘要: The nature of most environmental contaminants comes from chemical mixtures rather than individual chemicals. Most the existed mixture models are only valid for non-interactive toxicity. Therefore, we built two simple linear regression-based concentration addition (LCA) and independent action (LIA) that aim to predict combined toxicities interactive mixture. LCA model was between negative log-transformation experimental expected effect concentrations (CA), while LIA developed (IA). Twenty-four pesticide ionic liquid were used evaluate predictive abilities models. correlated well with observed responses 24 binary mixtures. values coefficient determination (R (2)) leave-one-out (LOO) cross-validated correlation (Q(2)) larger 0.99, which indicates high powers results showed allow accurately predicting synergism, additive effect, antagonism. proposed may serve as a useful tool in ecotoxicological assessment.