Penn researchers use AI to accelerate RNA drug development

Penn researchers are working to accelerate the development of drug delivery systems through a new artificial intelligence-driven research hub.

The United States National Science Foundation Artificial Intelligence-driven RNA Foundry facility — funded by a six-year, $18 million NSF grant — brings together researchers to design and test lipid nanoparticles that deliver RNA-based therapies. The program seeks to “transform how RNA is designed, synthesized and delivered” with AI and automation technology. 

Chemical and biomolecular engineering professor Daeyeon Lee, who also serves as AIRFoundry’s director, told The Daily Pennsylvanian the initiative is centered on expanding access to RNA technology across disciplines.

“The vision is very simple: democratize the RNA technology,” Lee said. “We’ve all seen the impact of RNA technology. During the pandemic, tens of millions of lives were saved by mRNA vaccines, which were invented here at Penn.”

Vaccines using mRNA were first proven effective by neurosurgery professor Katalin Karikó and vaccine research professor Drew Weissman, whose research helped inform the manufacturing of the COVID-19 vaccine.

Lee also explained that AIRFoundry seeks to reduce the technical barriers that currently limit broader uses of RNA tools.

“One of the challenges of RNA is that the entry barrier is fairly high,” he explained. “If you don’t know how to design RNA, synthesize RNA, or deliver RNA, it becomes very difficult to take advantage of the technology. So we wanted to lower that barrier.”

Lee added that the project’s long-term structure depends on building an “open platform” that allows researchers to “learn about RNA and how it can impact their research.”

“The hope is that researchers will share their results with our AI so that it continues to improve and make better recommendations for future users,” Lee said.

Third-year graduate student Andrew Hanna, who contributes research to the center, told the DP that AIRFoundry has helped establish a “shared institute for collaboration across lots of different groups.”

“There are a lot of different fields that come together in the work I’m doing, and the collaborators that I’ve been able to work through that have been facilitated by the Foundry have been really useful,” Hanna said.

Hanna helps design lipid nanoparticle formulation systems used to generate datasets for machine learning models. The goal of his work, he explained, is to speed up drug development by improving predictive systems and reducing trial-and-error testing.

“Ultimately, with enough data, the hope is that you can get a head start and actually be able to predict a drug without having to do nearly as much testing,” he said.

---