An alternative QSAR-based approach for predicting the bioconcentration factor for regulatory purposes

Area of Science:

• Environmental Chemistry
• Toxicology
• Computational Chemistry

Background:

• REACH and BPR regulations encourage non-animal testing for chemical risk assessment.
• Quantitative Structure-Activity Relationship (QSAR) models are gaining importance as alternatives to animal testing.
• Assessing Bioconcentration Factor (BCF) is crucial for regulatory compliance but is costly and time-consuming using traditional methods.

Purpose of the Study:

• To develop and validate QSAR models for predicting chemical Bioconcentration Factor (BCF).
• To provide a reliable, cost-effective, and ethical alternative to in vivo animal testing for BCF assessment.
• To support regulatory compliance under REACH and BPR frameworks.

Main Methods:

• Development of QSAR models using the ANTARES dataset, comprising verified experimental BCF data.
• Utilized a nine-descriptor model trained on 608 chemicals.
• Validated the model using separate validation (152 chemicals) and blind (76 chemicals) datasets.
• Implemented multi-step applicability domain assessment and safety margins for robustness.

Main Results:

• A highly predictive nine-descriptor QSAR model was developed.
• The model demonstrated satisfactory predictive power on external and blind datasets.
• The model's robustness was confirmed through rigorous validation strategies.
• The use of meaningful biokinetics descriptors ensures model interpretability.

Conclusions:

• The developed QSAR models offer a reliable alternative to in vivo BCF assays.
• These models can assist registrants in meeting regulatory requirements efficiently and ethically.
• The study supports the reduction of animal testing in chemical risk assessment.