AI startup aims to revolutionize drug discovery by simulating chemical interactions

A British startup, Potential Sciences, is leveraging artificial intelligence to accelerate the complex process of drug design by simulating molecular interactions rather than relying solely on laboratory testing. The company plans to build vast chemical datasets by modeling interactions between atoms, small molecules, and proteins using physics principles, aiming to train AI systems capable of identifying highly targeted drugs for specific ailments. Potential Sciences secured seed funding in the tens of millions of dollars from SovAI, the UK government’s AI venture capital fund, alongside Twin Path Ventures. Founder and CEO Hatem Helal noted the goal is to quickly eliminate ineffective molecules. While academics like Alan Fersht expressed optimism about the simulation approach, Professor Peter Wolynes of Rice University cautioned that current protein engineering techniques are still rudimentary. The startup's approach differs from many current labs that rely on the limited Protein Data Bank. Joséphine Kant, head of ventures at SovAI, highlighted that the physics-based angle sets Potential Sciences apart, granting them access to the UK’s national AI supercomputer. However, the company faces competition in a crowded field and must still address the significant hurdles of testing drug safety and efficacy before treatments reach the market.