Human beings have long lived in an uneasy race against the viruses that surround them, always reacting after a new threat has already arrived. Now researchers at the University of Cambridge have let artificial intelligence read the genetic codes of several coronaviruses and draft a single antigen meant to prepare the body for a whole family of them at once. The first people to receive it have shown that it is safe, and their immune response, though modest, has been read as a hopeful beginning. Whether this early promise becomes real protection will be decided in a larger study now underway, an
AI-Designed 'Super-Antigen' Vaccine Enters Human Trials, Targeting Whole Virus Families
A modest response at this stage is not a verdict on effectiveness
Mark
So what is actually new here? Vaccines already use antigens.
Mimi
The design. AI analyzed the genetic codes of several known coronaviruses and produced one antigen meant to cover the whole family, not a single virus.
Mark
And it's been tested on people?
Mimi
Yes, 39 people in the first trial, and it was confirmed safe. A second study with 200 people is underway.
Luke
Safe is confirmed. Protective is not. The report calls the immune response modest, and nothing here shows it prevents infection.
Mark
Why do the experts sound so excited, then?
Mimi
Andy Pollard, who wasn't involved, said people wouldn't have predicted these immune responses. Human immunity is far more complex than a mouse's, so any response in people matters.
Luke
Worth noting that the excitement is about the approach. The quotes about saving lives and game changers are about AI's potential, not about results from this vaccine.
Mark
What about the flu and Ebola plans?
Mimi
Those are research directions. A flu vaccine without annual updates, and one for Ebola strains that have no vaccine today.
Luke
Right, and no data on either is reported yet. Those are goals, not findings.
Mark
So the 200-person study is the one to watch.
Mimi
Exactly. It's built to measure how well the immune system is actually trained.
Der Puls
- A vaccine built around a single AI-designed antigen has moved from the lab into the first human volunteers, a step that carries both hope and scrutiny.
- The first trial in 39 people confirmed safety, but the immune response it produced was only modest, leaving open whether it can truly protect anyone.
- A 200-person study is now underway to measure how well the vaccine trains the immune system, the test that will separate promise from proof.
- Researchers are racing to widen the approach to seasonal flu and Ebola-strain viruses, aiming to end annual vaccine updates and cover threats that have no vaccine today.
- The field is watching closely, with outside experts calling the data fascinating and AI a possible game changer for faster, life-saving vaccine development.
Human beings have long lived in an uneasy race against the viruses that surround them, always reacting after a new threat has already arrived. Now researchers at the University of Cambridge have let artificial intelligence read the genetic codes of several coronaviruses and draft a single antigen meant to prepare the body for a whole family of them at once. The first people to receive it have shown that it is safe, and their immune response, though modest, has been read as a hopeful beginning. Whether this early promise becomes real protection will be decided in a larger study now underway, and the answer will tell us whether machines can help us get ahead of the next outbreak rather than only chase the last one.
A vaccine whose central ingredient was designed by artificial intelligence has now been given to people for the first time. The work comes from researchers at the University of Cambridge, and the early results appeared in the Journal of Infection.
The key piece is the antigen, the part that teaches the immune system what to look for. Instead of building around a single virus, the team had AI analyze the genetic codes of several known coronaviruses and shape them into one "super-antigen." The goal is to prepare the body to recognize a whole family of viruses, including ones that mutate or jump from animals to humans.
The first trial, involving 39 people, confirmed the vaccine is safe. Its immune response was modest but described as encouraging. A modest start is not a final verdict on effectiveness, and that question now moves to a larger study of 200 people already underway, which will measure more precisely how well the vaccine trains human immunity.
Outside experts were cautiously impressed. Prof Andy Pollard of the Oxford Vaccine Group, not involved in the work, called the data fascinating and said AI could become a game changer for vaccine development. The researchers, led in spirit by Prof Jonathan Heeney, hope to get ahead of outbreaks so people can be protected before new pandemics arrive.
The team is also extending the method to seasonal flu, where it could end the need for annual updates, and to Ebola strains that currently have no vaccine. The 200-person study will show whether that first flicker of immune response grows into real, useful protection.
A vaccine whose central ingredient was designed by artificial intelligence has now been given to people for the first time. The work comes from researchers at the University of Cambridge, and the early results were published in the Journal of Infection.
The ingredient in question is the antigen, the part of a vaccine that shows the immune system what to look for. Rather than building it around one virus, the team used AI to analyze the genetic codes of several known coronaviruses and design a single "super-antigen" from them. The aim is to train the body to recognize a whole family of viruses, including ones that mutate or cross from animals into humans.
The first human trial involved 39 people. According to the report, it confirmed that the vaccine is safe. The immune response it produced was modest, the findings noted, though described as highly encouraging. A modest response at this stage is not a verdict on effectiveness; that is the job of the next step.
A second, larger study of 200 people is now underway. Its purpose is to measure more precisely how well the vaccine trains the human immune system. Experts point out that human immune systems are far more complex than those of laboratory mice, which is why human trials are considered the real test.
Prof Jonathan Heeney of the University of Cambridge described the ambition in plain terms: the team is trying to get ahead of the curve, so that people could be protected against new outbreaks or pandemics before they arrive.
Outside voices were cautiously impressed. Prof Andy Pollard, director of the Oxford Vaccine Group, was not involved in the study. He told BBC News the data were fascinating and that people would not have predicted such immune responses could be generated. He added that AI looks set to be a game changer for vaccine research, with tools that could predict how the immune system will respond to a vaccine, make development much faster and save lives. Prof Marian Knight called the trial a pivotal leap forward in delivering broad, lasting viral protection.
The team is also looking beyond coronaviruses. The report says the method worked very well against rapidly changing viruses, and researchers are extending it to other threats. One target is seasonal flu, where such a vaccine could remove the need for annual updates. Another is viral hemorrhagic fevers: teams are working on a vaccine against Ebola strains for which no vaccine currently exists.
What to watch is the 200-person study. It will show whether the modest immune response seen in the first 39 volunteers grows into measurable, useful protection.
Bemerkenswerte Zitate
It's fascinating data and people wouldn't have predicted they'd be able to generate these immune responses.— Prof Andy Pollard, Oxford Vaccine Group, to BBC News
The team is trying to get ahead of the curve so people could be protected against new outbreaks or pandemics (paraphrased).— Prof Jonathan Heeney, University of Cambridge
Möchten Sie die ganze Geschichte?
Das Original lesen bei The News International ↗
Technology