Biomedical engineers at Duke University have developed a platform that combines automated lab methods with artificial intelligence to design nanoparticles for drug delivery.
The system helps researchers identify and optimise recipes for nanoparticles – tiny carriers that encapsulate drugs and deliver them to specific sites in the body.
In a proof of concept, it created nanoparticles able to deliver a difficult-to-encapsulate therapy for leukaemia and optimised the design of a second anti-cancer nanoparticle.
AI tools have already transformed drug discovery by predicting the properties of potential molecules.
But most platforms only support early-stage discovery, not later stages such as formulation and delivery, which are critical for safety and effectiveness.
Zilu Zhang is a PhD student in the lab of Daniel Reker, assistant professor of biomedical engineering.
Zhang said: “When you’re creating a nanoparticle, how well it works doesn’t just depend on the recipe, but also on the quantity of the various ingredients, including both the active drug and inactive materials.
“Existing AI platforms can only handle one or the other, which limits their overall effectiveness.”
Existing models also face challenges: some require very large datasets to function, while others struggle to distinguish between similar materials.
Many rely on fixed ratios, limiting their ability to optimise mixtures.
Reker and Zhang developed TuNa-AI to overcome these barriers. Using an automated liquid handling system, the team generated 1,275 distinct formulations combining therapeutic molecules and excipients – non-active substances such as colouring agents, preservatives and molecules that improve physical properties and absorption.
Zhang said: “By using robotics, we were able to combine many different ingredients in many different recipes very systematically.
“Our AI model was then able to look at that data for how different materials perform under different conditions and extrapolate that knowledge to select an optimised nanoparticle.”
The model achieved a 42.9 per cent increase in successful nanoparticle formation compared with standard methods.
As proof of concept, it formulated a nanoparticle that more effectively encapsulated venetoclax, a chemotherapy for leukaemia.
These nanoparticles showed improved solubility and halted leukaemia cell growth in the laboratory more effectively than the non-encapsulated drug.
In a second case study, TuNa-AI reduced the use of a potentially carcinogenic excipient by 75 per cent in another chemotherapy formulation, while maintaining efficacy and improving biodistribution in mouse models.
Zhang said: “We showed that TuNa-AI can be used not only to identify new nanoparticles but also optimise existing materials to make them safer.
Reker added: “AI can help us identify promising delivery molecules, but if you don’t mix them with the drug at a certain ratio, they won’t form a stable nanoparticle.
“If we can identify the optimal mixture ratios, then we can form the particles and maintain their stability.”
Beyond nanoparticle development, the team is collaborating with researchers and physicians both inside and outside Duke to expand the platform to other therapies.
Reker said: “This platform is a foundational step for designing and optimising nanoparticles for therapeutic applications.
“Now, we’re excited to look ahead and treat diseases by making existing and new therapies more effective and safer.”
