Artificial intelligence has been applied to design candidate drugs for antibiotic‑resistant infections and genetic diseases, but efforts to use it in the development of lipid nanoparticles (LNPs)—the delivery vehicles behind mRNA vaccines—have been limited. Designing LNPs is complex because each formulation combines multiple lipid components whose ratios determine how efficiently genetic material enters cells. The major obstacle has been the lack of large, systematic datasets that connect chemical composition to biological outcomes.
Now, engineers at the University of Pennsylvania have developed LIBRIS, short for “LIpid nanoparticle Batch production via Robotically Integrated Screening,” to bridge this gap. LIBRIS is an automated microfluidic platform capable of producing LNPs quickly and in large numbers, generating the data necessary to train predictive artificial intelligence models. “This could accelerate lipid nanoparticle development by as much as 100‑fold,” says Michael J. Mitchell, co‑senior author of the study published in ACS Nano. Because there are roughly 10¹⁵ possible formulations, expanding data generation is essential for exploring this enormous design space effectively.
LNP production typically involves creating ionizable lipids, formulating them into particles, and then testing their properties. While the first and last steps already yield extensive datasets, the formulation process itself has remained too slow and limited. “Actually formulating the nanoparticles is the bottleneck,” says Andrew Hanna, first author of the paper.
Search Antibodies Search Now Use our Antibody Search Tool to find the right antibody for your research. Filter
by Type, Application, Reactivity, Host, Clonality, Conjugate/Tag, and Isotype.
LIBRIS overcomes that bottleneck by combining microfluidic precision with robotic automation. The device looks like a miniature factory: tubes carrying lipid solutions feed into a glass microfluidic chip inside an aluminum housing, where the ingredients mix under precisely controlled conditions. A moving well plate collects each batch of particles, while rapid in‑chip cleaning allows continuous operation. The system’s parallel channels can create up to eight distinct formulations at once, producing approximately 1,000 per hour—around 100 times faster than conventional methods.
According to co‑senior author David Issadore, “If we can generate large, well‑defined libraries of LNPs, then we can start producing the data sets needed to identify patterns that can unlock the full potential of LNP‑based therapeutics.” By enabling faster, high‑precision formulation, LIBRIS could allow true rational design of nanoparticles—letting researchers specify desired properties in advance instead of relying on trial and error.