AI Finds New Monkeypox Target

FEATURED RESEARCH

AI Helps Scientists Identify a New Vaccine Target for Monkeypox and Future Outbreaks

A new study points to a surprising breakthrough in the race to develop better tools against mpox, a virus that infected more than 150,000 people during the 2022 global outbreak.

Using AlphaFold 3, researchers at UT Austin identified a single viral protein that can trigger strong neutralizing antibodies, something other researchers had never been able to figure out before.

A new way to find vaccine targets: Monkeypox virus (MPXV) carries dozens of proteins on its surface, and past attempts to guess at which one drives effective immunity was extremely challenging.

Traditional smallpox-based vaccines work, but they’re expensive to manufacture because they require whole weakened virus.

As a result, the investigators looked at alternatives. The researchers first collected 12 antibodies from people who had survived mpox or been vaccinated. They knew these antibodies neutralized the virus but not what they were binding to. That’s where AI entered the picture.

Using AlphaFold 3, the UT Austin team screened roughly 35 possible surface proteins. The model predicted (with unusually high confidence) that several of the potent antibodies targeted a protein called OPG153.

No one had considered this protein before and follow-up experiments confirmed the prediction.

What the experiments showed: When mice were injected with the AI-identified protein, they produced antibodies that neutralized MPXV. This makes OPG153 a promising candidate for a next-generation vaccine antigen or antibody therapy.

And because it’s just one protein, rather than an entire engineered virus, it could be faster, cheaper and safer to manufacture.

Why it matters: A single, easy-to-produce protein will result in mpox prevention. MPXV is closely related to the smallpox virus, so this approach could also strengthen biodefense efforts.

The team calls the method “reverse vaccinology”: start with survivors’ antibodies, let AI find the matching viral target, then design the vaccine around it.

It’s early, but this is the most precise target scientists have had for mpox. And it only surfaced because AI pointed them to a viral protein everyone had overlooked.

For more details: Full Article 

Brain Booster

Select the right answer! (See explanation below and source)

What Caught My Eye

PRODUCT PIPELINE

New Trial Data Strengthens The Case For AI In Identifying Viable Drug Candidates

AI drug discovery keeps making promises, but clinical proof has been thin. Recursion just gave this field a little more hope.

The company reported early trial data showing its AI-identified therapy, REC-4881, cut polyp growth in patients with familial adenomatous polyposis, a rare genetic condition that often pushes people toward colon removal.

Nine of eleven patients kept a durable reduction in polyp burden after stopping treatment.

The median drop hit fifty-three percent at twelve weeks post-therapy. For a disease where surgery has been the main safety net, this is a meaningful outcome.

Recursion calls this its first clinical validation moment. The drug was flagged by the company’s discovery engine, which screens biological patterns at scale.

Many groups in the space make similar claims, but few have matched them with patient data.

The trial started with a smaller readout in May that showed a forty-three percent reduction at thirteen weeks on treatment. The company now plans to expand the study and meet the FDA next year to map a registration path.

If this momentum holds, AI-driven discovery edges a little closer to proving its value where it counts.

For more details: Full Article

Top Funded Startups

Byte-Sized Break

📢 Other Happenings in Healthcare AI

• PathAI’s AIM-MASH AI tool just became the first AI system FDA-approved for use in MASH clinical trials, streamlining liver biopsy assessments to speed up drug development. [Link]

• Evaxion unveiled new preclinical data showing its off-the-shelf AML cancer vaccine, EVX-04, developed using its AI-Immunology platform, triggers strong immune responses and blocks tumor growth, with broad potential across hard-to-treat cancers. [Link]

• Nature’s deep dive on AI-designed drugs. [Link] Our insights on what makes an AI-designed drug. [Link]

Have a Great Weekend!

Trivia Answer: B) Penicillin

In 1928, Alexander Fleming returned from vacation to find that mold had killed bacteria in one of his culture dishes. That mold turned out to be Penicillium notatum, and it led to the discovery of penicillin, the first true antibiotic, which has saved millions of lives since.