Framework for In Silico Toxicity Screening of Novel Odorants

Area of Science:

• Environmental Chemistry
• Toxicology
• Computational Chemistry

Background:

• Chemical risk assessment often lacks experimental data for novel substances.
• Existing in silico models do not adequately predict inhalation toxicity for volatile chemicals in solution.

Purpose of the Study:

• To develop an in silico approach for estimating safe maximum solution concentrations of novel odorants.
• To utilize structure-based predictions for toxicology-based risk assessment.

Main Methods:

• Adapted decision trees from open-source mutagenicity and systemic toxicity models (Toxtree).
• Assigned Thresholds of Toxicologic Concern (TTC) based on predicted hazard classifications.
• Calculated maximum solution concentrations using predicted vapor pressure (MPBPWIN™) to ensure TTC is not exceeded upon inhalation exposure.

Main Results:

• The in silico approach successfully estimated safe solution concentrations for chemicals.
• Evaluated against a dataset of 143 chemicals, the method provided health-protective predictions for 98.6% of the test set.
• Demonstrated the utility of open-source models for predicting chemical safety.

Conclusions:

• The proposed in silico method enables the estimation of safe, toxicology-based maximum solution concentrations.
• This approach is valuable for assessing novel chemicals where experimental toxicity data is unavailable.
• Utilizes freely available software and models for accessible chemical risk assessment.