A research team at the University of Cambridge used artificial intelligence (AI) to design a so-called "superantigen" intended to provide protection across a broad range of coronaviruses. According to the researchers, this is the first time a key vaccine component has been entirely designed by AI and then tested in people.
The approach aims to train the immune system to recognise features common to many coronaviruses, reducing reliance on the specific surface proteins that viruses often change through mutation. The developers say the protection would therefore potentially cover not only known Covid variants but also coronaviruses circulating in animals that could spark future outbreaks.
Method: genetic data, AI analysis and antigen design
The team collected known genetic sequences of various coronaviruses from surveillance programme databases and used AI to analyse them. Based on this analysis, the algorithm designed an antigen intended to prepare the immune system against a whole virus family — even if individual viruses mutate or a new animal-to-human spillover occurs.
Jonathan Heeney, a professor at the University of Cambridge, described the approach as a fundamental shift in pandemic preparedness, with the goal of developing vaccines that protect against today’s pathogens and those that might cause the next outbreak.
Clinical trials and current findings
An initial human study enrolled 39 participants; the primary aim was to assess safety. Results published in the Journal of Infection reported that the immune effect so far is "moderate." The Cambridge team plans a second trial involving around 200 participants to obtain a clearer picture of how effectively the vaccine can prime the immune system.
Saul Faust, a professor at the University of Southampton who led part of the trial, said there are "clearly great opportunities" in AI-based vaccine design and described the findings as "extremely exciting."
Other targets: influenza, H5N1 and Ebola
The researchers are already applying the AI-based approach to other vaccine projects. They are working on a universal seasonal influenza vaccine that would not need annual updates to match circulating strains, and on an H5N1 avian influenza vaccine in case the currently bird-restricted virus evolves into sustained human-to-human transmission.
They are also exploring vaccines against viral haemorrhagic fevers, including different variants of the Ebola virus. The report notes that the current outbreak in the Democratic Republic of the Congo is caused by an Ebola strain for which no vaccine is yet available.
What remains to be seen
While the AI-designed antigen approach is promising, current results are preliminary and show only a moderate immune response. Larger trials and additional data are needed to determine whether AI-designed "superantigens" can deliver durable, broad protection. The researchers emphasise the potential of the technology to aid development of vaccines across multiple virus families, including influenza and haemorrhagic fever viruses.
