mRNA technology
Instructions written as RNA let the body make its own vaccine or drug protein, designed in days.
Open in the interactive tree →Katalin Kariko and Drew Weissman found around 2005 that modified RNA building blocks avoid inflammation, which made mRNA usable as a drug; they received the 2023 Nobel Prize in Physiology or Medicine. The UK authorized the first mRNA vaccine, Pfizer-BioNTech, on 2 December 2020, less than a year after the virus genome was published. The same platform can be reprogrammed within weeks, which opened vaccines, cancer therapy and gene-editing delivery.
As of October 2026
Moderna's mRNA RSV vaccine was authorized in the US in May 2024, a combined flu and COVID vaccine in the European Union in February 2026 and an mRNA flu vaccine in the US in August 2026 for people over 50. On 19 August 2026 Merck and Moderna reported that intismeran autogene, an individualized mRNA cancer vaccine, met its primary endpoint of recurrence-free survival and its key secondary endpoint of distant metastasis-free survival in the phase 3 INTerpath-001 trial (1,137 patients with resected stage IIB-IV melanoma, given with Keytruda), the first positive phase 3 result for an mRNA cancer vaccine. Full data, including effect size, were not yet public in the announcement.
Open steps
- Delivery beyond liver and muscle High AI leverageLipid nanoparticles mostly reach the liver and muscle; reaching lung, immune organs, brain or tumours safely would widen mRNA to many more diseases.
- Targets for personal cancer vaccines High AI leverageIndividualised mRNA vaccines must pick the few mutations that immune cells really see; predicting presented, immunogenic neoantigens is still weak.
- Longer-lasting, cheaper mRNA High AI leverageSequence and chemistry set how long mRNA lasts and how much protein it makes; better designs and freezer-free storage would cut dose, cost and cold chain.
- Self-amplifying mRNA beyond COVID Medium AI leverageSelf-amplifying RNA copies itself in cells and can cut doses, but long constructs, side effects and manufacturing must be shown for uses beyond COVID.
Where AI could help
Medium AI leverage. AI tunes sequences and lipid carriers, but trials, manufacturing and delivery beyond liver and muscle set the pace.
- Optimizing mRNA sequences for stability and protein output
- Predicting ionizable lipids for organ-specific delivery
- Selecting tumor mutations for individual cancer vaccines
- Fast antigen redesign when a virus mutates
Shown so far
- In May 2023 Nature published LinearDesign, an algorithm for mRNA sequences that raised mRNA half-life and gave up to 128-fold higher antibody titres in mice than codon optimization. source
- In 2024 Nature Communications published AGILE, a deep-learning platform that screened about 12,000 virtual ionizable lipids and found one delivering mRNA to muscle 7.8 times better than a standard lipid in preclinical tests. source
Prerequisites
- Vaccination1796
- Randomized controlled trial1948Randomized phase 3 trials proved the vaccines worked
- Genetic engineering1973
- RNA interference1998Lipid nanoparticles proven for the first RNA-interference drug also carry mRNA vaccines
- Genome sequencing2003
- Cryo-EM Resolution Revolution2017Cryo-EM confirmed the stabilised prefusion spike used as the 2020 vaccine antigen
- COVID-19 pandemic2020mRNA vaccines were first used at scale in the COVID-19 response