A machine learning approach to predict cellular uptake of pBAE polyplexes.

Area of Science:

• Biomaterials Science
• Nanotechnology
• Gene Therapy

Background:

• Gene delivery systems are crucial for treating diseases but face limitations in cellular uptake efficiency.
• Poly β-amino esters (pBAEs) are effective polymer vectors forming polyplexes for oligonucleotide delivery.
• Optimizing pBAE formulations for specific cell types currently relies on time-consuming trial-and-error methods.

Purpose of the Study:

• To develop and validate machine learning (ML) models for predicting poly β-amino ester (pBAE) polyplex cellular internalisation.
• To identify key features influencing pBAE nanoparticle uptake and transfection efficiency across different cell lines.
• To establish an in silico screening tool for accelerating the optimization of gene delivery formulations.

Main Methods:

• Fabrication of a library of poly β-amino ester (pBAE) nanoparticles.
• Assessment of nanoparticle uptake in four distinct cell lines.
• Training and evaluation of various machine learning models, including gradient-boosted trees and neural networks.
• Interpretation of the best-performing model using SHapley Additive exPlanations (SHAP).

Main Results:

• Machine learning models, particularly gradient-boosted trees and neural networks, were successfully trained to predict pBAE polyplex uptake.
• The developed ML models demonstrated effectiveness in learning complex, non-linear relationships within the data.
• SHapley Additive exPlanations provided insights into the critical features driving cellular internalisation predictions.

Conclusions:

• Machine learning offers a powerful in silico tool to predict and optimize poly β-amino ester (pBAE) nanoparticle delivery systems.
• This predictive capability can significantly reduce the experimental burden associated with finding optimal gene delivery formulations for new cell types.
• The findings pave the way for accelerated development of efficient and targeted gene therapies.