AI Startup Skips Lab Work To Find New Drugs

Potential Sciences, a newly launched British startup, is leveraging artificial intelligence to accelerate the traditionally sluggish and unpredictable process of drug design. By bypassing extensive laboratory experimentation, the company aims to streamline the identification of highly targeted therapies for specific diseases. The approach relies on generating large-scale chemical datasets through physics-based simulations that model interactions ranging from atoms to complex proteins. This data serves as the foundation for training AI models capable of predicting how potential drugs will behave within human biology, effectively filtering out ineffective compounds before they reach clinical stages.

“We want molecules that are never going to work in the clinic to fail fast,” Hatem Helal, Potential Science’s founder and CEO, tells WIRED. “That means having better predictors of how they will behave as they interact with human biology.”

The company officially launched on November 5, securing seed funding in the tens of millions of dollars from SovAI, the UK government’s AI venture capital fund, alongside Twin Path Ventures and other investors. While the founders have declined to disclose the exact valuation or total amount raised, the financial backing underscores significant confidence in their methodology. If successful, this strategy could drastically reduce the costs associated with drug development, making it economically viable for pharmaceutical companies to investigate treatments for rare conditions that have historically been neglected due to high experimental expenses.

Helal envisions a future where medicine becomes more personalized. “We want to make it possible to have more bespoke treatments for diseases,this idea that medicines can be one-off molecular combinations that are tailored to your unique human biology,” he says.

However, the feasibility of this ambitious goal remains a subject of debate among experts. The core challenge lies in whether Potential Sciences can accurately predict how well drug candidates bind to disease-related proteins, a task that requires significant advancements in current computational capabilities. Peter Wolynes, a professor of chemistry at Rice University, offered a measured perspective on the startup’s prospects. “I’m sure it will be an improvement,” Wolynes states. Yet, he cautioned that the current state of the field lags behind historical benchmarks. “But currently, ‘the engineering is at the level of the Romans, not of the Renaissance, in terms of how protein engineering is done,’ he says.”

Despite these technical hurdles, the broader industry remains optimistic about the transformative power of AI in science. Prominent figures such as Demis Hassabis, who was awarded the Nobel Prize for Chemistry for his work at Google DeepMind, have long argued that artificial intelligence will drive momentous scientific breakthroughs, similar to its recent impact on mathematics. Startups like Potential Sciences represent a growing belief that drug discovery may be one of the first major scientific frontiers to undergo such a revolution, fundamentally changing how new medicines are developed and delivered.