Researchers at MIT published results in Nature Biotechnology on September 28, 2026, showing that an AI-assisted method can stabilize experimental mRNA-carrying lipid nanoparticles at temperatures that degrade many current formulations. Lipid nanoparticles are the protective shells that deliver fragile mRNA into human cells. The results come from animal studies, not human trials, and no regulatory body has approved changes to existing vaccine storage requirements.
How the Algorithm Works
An iterative machine-learning approach compressed months of conventional screening into several weeks.
The team’s machine-learning algorithm screened nearly 50 FDA-approved excipients, the inactive ingredients such as sugars, salts, and polymers that surround the nanoparticles. From those, it proposed ratios for the five most promising candidates.
Researchers tested those predictions, fed the results back into the model, and repeated the cycle until a stable formulation emerged. That process took several weeks rather than the months of trial-and-error that would otherwise have been required.
Key performance results from the study, all based on preclinical data:
- Experimental vaccine particles remained stable for up to one year at room temperature after vacuum drying
- Stability held for two months at 37°C (approximately 98.6°F)
- Formulations were tested on Covid-19 mRNA antigens in Moderna-like lipid nanoparticles; the algorithm also stabilized a Pfizer-like formulation
- Mice vaccinated after elevated-temperature storage showed immune responses comparable to those from standard refrigerated vaccines
- Solid microneedle patches containing a SARS-CoV-2 antigen produced similar immune responses in mice
Ana Jaklenec, principal investigator at MIT’s Koch Institute for Integrative Cancer Research and a senior author on the paper, described what sets the approach apart: “The real beauty of this algorithm is that we can use it with small data sets.”
Why Cold Chains Break Vaccine Access
Refrigeration dependency creates hard logistical barriers that this research aims to reduce.
Many current mRNA vaccine formulations require refrigerated or frozen cold-chain storage, and mRNA degrades without the protection that lipid nanoparticles provide. Regions without reliable power grids or refrigeration infrastructure cannot maintain that chain, and that is where distribution fails.
Reducing that dependency could simplify shipping, lower distribution costs, and expand access in low-resource settings. The same formulation strategy, the researchers reported, could extend beyond Covid-19 to cancer immunotherapies, other infectious-disease vaccines, and RNA therapeutics more broadly.
The project was led by graduate student Jinbi Tian and postdoctoral researcher Khanh Tran, with senior authorship from Jaklenec and Robert Langer, the David H. Koch Institute Professor. MIT’s Computer Science and Artificial Intelligence Laboratory, including Mina Konaković Luković, assistant professor of electrical engineering and computer science, also contributed. The Gates Foundation provided partial funding.
What Comes Next
Several mandatory steps separate this preclinical result from any approved change to vaccine storage.
Between this laboratory result and a pharmacy shelf sit several mandatory steps. Those include human clinical trials, large-scale manufacturing validation, long-term stability studies across different mRNA payloads, and full regulatory review.
If the approach clears those hurdles, it could meaningfully change how mRNA-based medicines reach underserved populations. For regions where the cold chain has always been the weakest link, that shift would be significant.
