The University of Cambridge has announced the start of the first clinical trial of a vaccine whose key component was developed entirely by artificial intelligence. AI has already been used to create drugs and individual vaccine components, and some drugs developed with its help have even reached clinical trials. However, until now it had been only one of the researchers’ tools.
In the Cambridge work, the algorithm independently analyzed genetic data from thousands of sarbecoviruses — a family that includes the pathogens behind SARS, COVID-19 and many bat coronaviruses — and selected the most stable regions of the viruses to form the basis of the vaccine.
The main feature of the development is its universal approach. Instead of protecting against one specific viral variant, the vaccine is meant to teach the immune system to recognize nearly unchanged elements shared by a large number of related coronaviruses. If the concept is confirmed in further studies, it could eventually make it possible to create vaccines that are less dependent on the emergence of new mutations.
The new vaccine is a DNA vaccine, which is usually easier to store and transport than mRNA-based products. It can also be administered without a needle — using a jet injector that delivers the liquid through the skin at high pressure.
Early results showed that the vaccine was well tolerated and triggered antibodies capable of recognizing different sarbecoviruses. However, the researchers stress that this is only an initial stage. They still need to assess how long the immune protection lasts, whether booster shots are necessary, and whether the vaccine can prevent infection in larger clinical studies.
