Broad COVID vaccine designed with AI tested in humans for the first time

An experimental DNA vaccine against a broad family of coronaviruses has been found safe in a first trial involving 39 volunteers. Researchers hope the approach could lead to broad-spectrum vaccines against pandemic threats in the future.

COVID-19 illustration: depositphotos.com
COVID-19 Illustration: depositphotos.com

Researchers from the University of Cambridge and the spin-out company DIOSynVax report a first step in developing a new generation of vaccines: an experimental vaccine whose main active ingredient was designed using artificial intelligence and computer simulations, and was tested for the first time in humans.

The vaccine, called pEVAC-PS, is designed against the cerebrospinal virus family, a subgroup of coronaviruses that includes SARS-CoV-2, the causative agent of the coronavirus pandemic, the original SARS virus, as well as related viruses found in bats that could potentially jump to humans in the future. The long-term goal is to develop a broad-spectrum vaccine that can protect not just against a specific strain that has already broken out, but against an entire family of related viruses.

The first human trial, published in the Journal of Infection, involved 39 healthy volunteers aged 18-50 at two clinical centres in the UK, in Cambridge and Southampton. Participants received the vaccine in increasing doses. The researchers said the vaccine was safe and well-tolerated, with no significant side effects. It also elicited measurable immune responses against several cervicoviruses, but the magnitude of the response was modest, so it is still early days.

Don't chase every variant.

Many vaccines work by presenting an antigen – a component of the virus that the immune system learns to recognize. The problem is that viruses change. When a vaccine is designed against a particular variant, it may lose some of its effectiveness as the virus accumulates mutations. That’s why flu vaccines are updated every year, and coronavirus vaccines have also been updated repeatedly since the beginning of the pandemic.

The new approach attempts to circumvent some of this problem. Instead of choosing an antigen from a single strain of virus, the researchers used genetic information from a wide range of cerebrospinal viruses collected in global surveillance databases. Using machine learning, they looked for conserved regions—features that recur across a broad family of viruses and tend to change less. From these regions, a “superantigen” was designed that would direct the immune system to common features rather than just one version of the virus.

Professor Jonathan Heaney of the University of Cambridge, who led the scientific aspect of the study, described it as a shift from a reactive approach to one that tries to be prepared in advance. That is, not waiting for the next virus and then starting to develop a new vaccine, but building vaccines in advance that cover families of viruses with pandemic potential.

DNA vaccine and no needle

The experimental vaccine is not an mRNA vaccine, of the type that was made public during the coronavirus pandemic, but a DNA vaccine. DNA vaccines include a genetic segment that instructs cells in the body to produce the desired antigen, thus triggering an immune response. One reason for the interest in such vaccines is their relative stability, which could make them easier to store and distribute, especially in countries with limited refrigeration infrastructure.

In the current trial, the vaccine is also being administered without a needle, using a device that creates a tiny jet of liquid and injects it into the skin. This delivery may make large vaccination campaigns easier, reduce dependence on syringes, and also suit people who are afraid of injections. However, both the delivery method and the vaccine itself are still in the testing stages.

The technology is not necessarily limited to COVID-19. Cambridge notes that the design principle could also be used in the future against other families of viruses, such as influenza or hemorrhagic fever viruses, including Ebola. The idea is to focus on conserved characteristics of a family of viruses, and develop a vaccine that will remain relevant even as viruses continue to mutate.

Great promise, but not yet a vaccine ready

Despite the enthusiasm, it is important to emphasize what the study has not yet proven. The trial is primarily designed to test safety, tolerability, and initial immune response. It is not designed to show that the vaccine prevents infection, severe illness, or person-to-person transmission of the virus. It is also unclear how long the immune response will last, whether booster doses will be needed, and how effective the vaccine will be in older populations or people with underlying medical conditions.

The next step will be a Phase 2 trial in a larger, more diverse population to see if the vaccine elicits a strong and broad enough immune response. Only after larger trials will we know if this is a platform that could change how we prepare for future pandemics.

Still, the very move from animal testing to the first human trial is an important milestone. If the approach holds up to future tests, it could add a new class of vaccines to the public health toolbox: not just vaccines against a virus that has already spread, but vaccines that are designed to protect against future viral threats.

Sources and image

Scientific article:
A phase I, needle free, dose escalation clinical trial of pEVAC-PS, a candidate pan-Sarbecovirus Vaccine
Journal: Journal of Infection, 2026;92(6):106759
DOI: 10.1016/j.jinf.2026.106759

Short FAQ:

What's new in this vaccine?
Its main active ingredient was designed using artificial intelligence and computer simulations, with the aim of identifying features common to a broad family of viruses.

Is the vaccine already approved for use?
No. This is an experimental vaccine that is being tested in the first phase in humans. It is not yet an approved vaccine for the public.

What did the first experiment find?
The trial found that the vaccine was safe and well-tolerated in 39 volunteers, and that it elicited certain immune responses against several cerebrospinal viruses.

Has it been proven to prevent coronavirus or future pandemics?
No. The trial is not intended to prove prevention of infection or disease. For that, larger and longer trials will be required.

More of the topic in Hayadan:

Leave a Reply

Email will not be published. Required fields are marked *

This site uses Akismet to filter spam comments. More details about how the information from your response will be processed.