Multiple toxicity endpoint-structure relationships for substituted phenols and anilines.

Area of Science:

• Environmental Toxicology
• Computational Chemistry
• Ecotoxicology

Background:

• Quantitative structure-activity relationship (QSAR) and quantitative structure-toxicity relationship (QSTR) models are crucial for predicting chemical toxicity.
• Predicting multiple toxicity endpoints for aquatic organisms like Chlorella vulgaris is essential for environmental risk assessment.
• Substituted phenols and anilines are common environmental contaminants whose toxicity needs reliable prediction.

Purpose of the Study:

• To develop QSTR models with a consistent mathematical structure for predicting various toxicity endpoints.
• To utilize norm index descriptors for modeling the toxicity of substituted phenols and anilines towards Chlorella vulgaris.
• To assess the predictive accuracy, robustness, and reliability of the developed QSTR models.

Main Methods:

• Development of QSTR models based on norm index descriptors.
• Prediction of four aquatic toxicity endpoints: IC50, IC20, LOEC, and NOEC.
• Validation using leave-one-out cross-validation, Y-randomized validation, and application domain analysis.

Main Results:

• The developed QSTR models demonstrated satisfactory predictive performance for all four toxicity endpoints.
• High squared correlation coefficients (R²) were achieved, indicating good model fit.
• Validation analyses confirmed the accuracy, robustness, and reliability of the models for predicting chemical toxicity.

Conclusions:

• QSTR models with the same mathematical structure can successfully predict multiple toxicity endpoints.
• Norm index descriptors are effective for developing reliable QSTR models for aquatic toxicity.
• The findings support the use of QSTR for environmental risk assessment of chemicals towards Chlorella vulgaris.