AI-guided optimization enables thermostable mRNA vaccines

Summary

A Nature Biotechnology research briefing describes AGENT, an AI-driven framework that combines high-throughput experimentation with Bayesian optimization to rapidly engineer thermostable, solid-state mRNA vaccine formulations. The formulations retained full bioactivity after storage at 37 °C for two months. In tests reported in the briefing, they produced immune responses that were non-inferior to those generated by fresh vaccines. The approach also enabled delivery without a conventional cold chain through microneedle patches. This delivery was demonstrated in rodents and nonhuman primates, indicating that the framework was evaluated across more than one animal model. The briefing presents the work as a way to address the storage and distribution constraints associated with mRNA vaccines, while the full research article is available through subscription access.