AI-designed proteins race to revolutionize snakebite treatments

There is an abundance of industries and fields that have implemented artificial intelligence (AI) to bolster their work. The latest use for the technology has now enabled the creation of custom-designed proteins that neutralize the deadly toxins found in snake venom. This breakthrough could revolutionize the treatment of snakebites, which currently claim around 100,000 lives annually and cause numerous disabilities, particularly in impoverished regions. These AI-designed proteins promise a new generation of therapies, addressing the limitations of century-old antivenom techniques.

Snakebite treatment has remained largely unchanged for over 100 years, relying on antivenoms derived from antibodies found in the blood serum of horses or sheep immunized with snake venom. These treatments are often unsafe, vary in effectiveness, and, as an article by the Nature publication highlights, require trained healthcare staff to administer them—posing significant barriers to accessibility in remote or under-resourced areas. Recognizing the urgent need for innovation, the World Health Organization (WHO) has designated snakebite as a high-priority neglected tropical disease.

The groundbreaking study, published in Nature journal on January 15, demonstrates how machine learning has transformed protein design. RFdiffusion, a computational tool developed by David Baker’s lab at the University of Washington, applies AI to rapidly design small proteins that bind tightly to target molecules. Inspired by generative AI tools like DALL-E, RFdiffusion can accomplish in seconds what previously took months or years—or was deemed impossible.

Initially developed to address conditions such as cancer and autoimmune diseases, the tool caught the attention of biochemist Susana Vázquez Torres, who saw its potential for neglected diseases like snakebite. Snake venom contains multiple toxins, including those that paralyze nerves, damage muscles, and destroy tissues. Vázquez Torres used RFdiffusion to design “mini-binders” that attach to and neutralize these toxins with remarkable precision.

Promising Experimental Results

The research team screened a few dozen AI-designed proteins for each venom component, identifying mini-binders that neutralized toxins effectively in lab tests. In experiments with mice, these proteins provided complete protection even against lethal doses of venom. When the mini-binders were injected 15 minutes after exposure to the venom, the mice still survived, simulating real-world scenarios where treatment might not be immediate. These results mark a significant milestone in snakebite research.

AI-designed proteins offer several advantages over traditional antivenoms. Mini-binders are extremely stable, potentially eliminating the need for refrigeration—a critical benefit for use in remote or resource-limited areas. They can also be produced at low cost using bacteria in industrial fermenters, making them more affordable than existing treatments. Furthermore, these synthetic proteins are highly specific, which could reduce side effects and improve efficacy.

However, the study focused on only a subset of venom components, such as neurotransmitter-targeting toxins. For a comprehensive antivenom, additional mini-binders will need to target other venom elements, such as phospholipases, which cause extensive tissue damage. Researchers envision a cocktail of mini-binders tailored to the venom profiles of specific regions to maximize their effectiveness.

While the results are promising, it may take years to bring these antivenoms to clinics. Financial hurdles remain a significant barrier. Unlike diseases such as cancer, which attract substantial investment, neglected tropical diseases like snakebite struggle to secure funding. Despite raising $1 billion for other protein design applications, Baker’s team faces a steeper climb to support treatments for diseases affecting the developing world.

Broader Implications

This breakthrough highlights how AI can accelerate solutions to global health challenges. Beyond snakebite treatments, AI-driven protein design is poised to tackle a range of neglected diseases. The research also underscores the importance of addressing funding disparities to ensure life-saving technologies benefit underserved populations.

As researchers refine these AI-designed antivenoms and navigate the path to clinical application, their work could fundamentally transform how snakebites are treated—potentially saving thousands of lives each year while reshaping the role of AI in combating neglected diseases.

By Nazrin Sadigova