Machine learning integrated with in vitro experiments for study of drug release from PLGA nanoparticles

Area of Science:

• Biomaterials Science
• Pharmaceutical Sciences
• Computational Biology

Background:

• Poly lactic-co-glycolic acid (PLGA) micro-/nano-particles are widely used for drug delivery.
• Understanding and predicting drug release kinetics from these particles is crucial for effective therapeutic outcomes.
• Existing models often require extensive experimental data and may not capture complex interdependencies.

Purpose of the Study:

• To investigate drug delivery from PLGA micro-/nano-particles using advanced machine learning algorithms.
• To identify and quantify the influence of key parameters (drug solubility, molecular weight, particle size, pH) on drug release profiles.
• To utilize machine learning insights for guiding the design of new in vitro drug release experiments.

Main Methods:

• Analysis of experimental data from approximately 50 published papers.
• Application of machine learning algorithms: linear regression, principal component analysis, Gaussian process regression, and artificial neural networks.
• Design and execution of new in vitro experiments informed by machine learning predictions.

Main Results:

• Machine learning models successfully identified significant correlations between drug properties, particle characteristics, environmental pH, and drug release rates.
• The predictive power of the machine learning algorithms was validated against experimental data.
• New in vitro experiments, guided by the ML analysis, showed results consistent with the model predictions.

Conclusions:

• Machine learning provides a powerful tool for understanding and predicting drug release from PLGA micro-/nano-particles.
• Key factors such as drug solubility, molecular weight, particle size, and pH significantly impact drug release kinetics.
• The integration of machine learning with experimental validation offers a pathway for optimizing drug delivery system design.