AI-engineered Vaccine Approach Could End Cycle of Variant Chasing

By Deborah Borfitz 

August 12, 2026 | A spin-out company of the University of Cambridge is developing a pipeline technology to serve as the computer chip for a whole new class of vaccines covering a family or group of viruses targeted at different stages of their lifecycle. DIOSynVax, short for Digital Immune-Optimized Synthetic vaccines, brings together genomics, synthetic biology, artificial intelligence (AI), and immunology to design broadly protective vaccine antigens that are compatible with most vaccine delivery systems, according to Jonathan Heeney, company cofounder and professor of comparative pathology at the university.   

Work in this direction followed the devastating 2014-2016 Ebola outbreak in West Africa that infected more than 28,600 people, killing 40% of them. Heeney and DIOSynVax cofounder Rebecca Kinsley subsequently led efforts to create a pan-hemorrhagic fever vaccine targeting a group of viruses, including Sudan, Marburg, and Lassa fever viruses, which present with the same clinical features as Ebola virus disease. 

These diseases share overlapping, sometimes generalized clinical features and the absence of specific tests in the field where cases occur cause diagnostic delays, often leading to an out-of-control epidemic before medical expertise arrives, says Heeney. DIOSynVax’s portfolio of vaccines has over the years expanded to include many viral families of global public health concern, including seasonal and pandemic influenza vaccines, coronaviruses (e.g., MERS, SARS, and SARS-CoV-2), and highly deadly zoonotic henipaviruses.  

Looking ahead, Heeney believes the company's technology could eventually become a standard part of preventive healthcare, supporting the development of vaccines against a wide range of infectious diseases, from endemic respiratory viruses to future pandemic threats. "To be clear, this is not a single vaccine intended to protect against all infectious diseases,” says Heeney. “Rather, the approach involves designing ... vaccines against specific viral families or pathogen groups, each intended for a particular clinical indication." 

A newly published phase 1 trial of a pan-sarbecovirus vaccine, designed to provide protection against the larger group SARS, SARS-CoV-2 (COVID), and related viruses, serves as an early proof of concept for the technology (Journal of Infection, DOI: 10.1016/j.jinf.2026.106759). The DNA vaccine, in this case administered through a jet of air targeting immune cells in layers within the skin, was shown to be safe, and participants developed measurable responses to the intended broad virus targets. 

Trial planning had concluded before the time Moderna and Pfizer gained regulatory approval for their messenger RNA COVID vaccines, and manufacturing for those trials took much longer, Heeney says. The vaccine of DIOSynVax differed from the original COVID-19 vaccines based on the Wuhan strain, which required regular updates to match new variants that were evolving faster than the first-to-market vaccines could be updated. 

The current priority for DIOSynVx is a seasonal flu vaccine, an annual inoculation given to hundreds of millions of people around the world. Influenza can be a fatal disease “that nobody wants next season for themselves or their elderly mother to acquire if there’s a good vaccine to prevent it,” says Heeney.  

Unlike most other seasonal influenza vaccines that contain three virus isolates—influenza A strains H1N1 and H3N2 plus one influenza B strain, chosen to match the variants predicted to circulate next season—the one under development by DIOSynVax is a “supra-seasonal” influenza vaccine offering protection against the huge spectrum of variants that an individual could potentially encounter over the course of several influenza seasons.  

“We’ve done all the pre-regulatory animal work and right now we’re hoping to be able to start manufacturing early next year,” says Heeney. Clinical trials of the flu vaccine can begin as soon as the funding support can be secured.  

Dialogue is ongoing with regulatory agencies in North America, Europe, and the U.K., Heeney reports. “We’re talking to all of them and hopefully we’ll soon have a clear path on how we move this next generation of vaccines into safe clinical practice as fast as possible.” 

Many Delivery Trucks 

Heeney is reticent of the “universal vaccine” term, which the DIOSynVx approach is sometimes called, because it can have a lot of different interpretations. “A single universal vaccine ... is really very much sci-fi stuff,” he says, as opposed to the company’s strategy offering broad protection against several virus subtypes. 

Vaccines themselves are delivered as a formulation that includes a vector, the contents of which are known as the antigen or antigens, Heeney explains. These are the target-specific payloads he likens to a package in a delivery truck, and that vehicle can be RNA, DNA, viral vector, live-attenuated, or inactivated vaccines or proteins in an adjuvant. They can all ferry antigens to immune cells inside the body to provide defensive immune responses. 

The original COVID-19 vaccines were based on one viral isolate and one target protein, in that case the coronavirus spike protein of the early clinical isolate identified as the Wuhan-1 strain. But by the time the first vaccine was manufactured, distributed, and being put in arms, almost a year had gone by, and the virus had already undergone mutations, says Heeney. So, people who were vaccinated but got infected with a new variant still got sick even though many lives were saved, inadvertently fueling some skepticism that eventually grew into anti-vaccine sentiment. 

“The fact is these viruses are always on the move and variants developed in the population, requiring the need to reformulate new vaccines with variant spikes,” he continues. That meant repeated booster shots for the population almost every year, as is currently done to protect people from seasonal influenza viruses. “In this context, we’re constantly behind the curve ... always chasing the variants.” 

This gets back to research underway at DIOSynVax to find “a single bullet” (supra-seasonal vaccine) for the whole group of flu viruses. The standard trivalent variety is not in any sense universal but based on three selected viruses from different subtypes (H1N1, H3N2, and influenza B strain) belonging to the same large Influenza family of viruses, Heeney says. Each virus is distinct and immunologically different using current vaccine technology that doesn’t induce cross-protection from different flu strains. 

Much like the combination therapy used to treat cancer and HIV, DIOSynVax is employing a comprehensive strategy—in this case, instructing the immune system to target different stages of the viral lifecycle to induce an immune response blocking several crucial steps in its replication cycle. AI is used to analyze globally sourced genetic sequence data, identifying stable and conserved features across entire virus families to computationally design synthetic and “antigenically unique” vaccine antigens, he explains. “What’s great about this technology is that we can plug and play [these vaccine antigen payloads] into any vaccine [delivery truck] that is best suited for a disease indication or large-scale manufacturing.”  

A supra-seasonal flu vaccine would offer protection “from all of the different potential variants that everybody brings together and shares over the Christmas table,” Heeney says. “That is a huge advance.”