Rational Design of Safer Inorganic Nanoparticles via Mechanistic Modeling-Informed Machine Learning.

Area of Science:

• Nanotechnology
• Toxicology
• Computational Biology

Background:

• Assessing inorganic nanoparticle (NP) safety is crucial for clinical applications.
• Existing methods for NP toxicity evaluation are time-consuming and resource-intensive.
• Predictive models are needed to streamline the safety assessment of novel nanomaterials.

Purpose of the Study:

• To develop a machine learning (ML) framework for predicting the in vitro and in vivo toxicity of inorganic NPs.
• To identify key physicochemical properties and experimental conditions influencing NP toxicity.
• To integrate physiologically based pharmacokinetic (PBPK) modeling for organ-specific in vivo toxicity predictions.

Main Methods:

• Trained and validated binary classification ML models using a curated in vitro cytotoxicity dataset.
• Performed explainability analysis to determine NP toxicity determinants and structure-toxicity relationships.
• Integrated a PBPK model into the ML pipeline to estimate organ-specific NP exposure for in vivo predictions.

Main Results:

• The ML framework accurately predicted in vitro NP toxicity across diverse inorganic NPs.
• Explainability analysis revealed critical factors governing NP-induced toxicity.
• PBPK-informed ML models demonstrated robust predictions of organ-specific nanotoxicity.

Conclusions:

• The developed PBPK-informed ML framework offers a streamlined approach for NP safety assessment.
• This approach facilitates the rational design of safer inorganic NPs.
• The framework has the potential to expedite the clinical translation of nanomaterials.