Modeling uptake of nanoparticles in multiple human cells using structure-activity relationships and intercellular...

Area of Science:

• Nanotechnology
• Computational Chemistry
• Toxicology

Background:

• Nanoparticle (NP) cellular uptake is crucial for biomedical applications but experimentally challenging.
• Predictive models are needed to streamline the screening of functionalized NPs for targeted delivery.

Purpose of the Study:

• To develop quantitative structure-activity relationship (QSAR) and quantitative activity-activity relationship (QAAR) models for predicting NP cellular uptake and viability.
• To identify key structural features influencing NP cellular permeability.

Main Methods:

• Developed decision treeboost QSAR models for predicting NP uptake in five human cell types.
• Created QAAR models to correlate cellular viability across different cell types.
• Defined applicability domains using the leverage method.

Main Results:

• QSAR models achieved high accuracy (R² > 0.914 in test sets) and cross-validation (Q² between 0.627-0.926).
• Models demonstrated low error (RMSE < 0.11, MAE < 0.09) and outperformed previous predictive methods.
• Identified NP structural features influencing cellular permeability and developed cross-cell type prediction models.

Conclusions:

• The developed QSAR and QAAR models offer reliable tools for predicting NP cellular uptake and viability.
• These models can accelerate the screening of novel NPs for specific cell affinities in biomedical research.
• The study provides a framework for in silico assessment of NPs, reducing experimental costs and time.