Skip to main content
CRS 2024 Annual Meeting and Exposition Main banner
Final Program


CRS Abstract Guide

Session: Technical Session V

Artificial Intelligence and Predictive Models in Pharmaceutical Technologies

Thursday, July 11, 2024
9:00 AM - 11:00 AM CET
Location: Magenta B

Sponsored By 
Machine learning to predict extracellular vesicle properties for drug delivery applications

Extracellular vesicles (EVs) are lipid-bound nanoparticles produced by all cell types. EVs carry small molecule, nucleic acid, and protein cargoes to facilitate long-distance cellular communication. Given their innate targeting properties, barrier crossing abilities, and lower immunogenicity, compared to synthetic lipid nanoparticle formulations, EVs represent a promising therapeutic modality. However, for clinical translation to become a reality, we must improve our ability to control, produce, and isolate EVs. Here, we demonstrate that tuning culture properties allows us to control EV properties. Further, by employing machine learning algorithms we can predict the parameters within the design space that result in EVs of a desired size, surface charge, and concentration. Our work has important implications for manufacturing EV-based therapeutics by eliminating the trial-and-error workflow used to selectively tune EV characteristics.