Potential Sciences Launches AI Drug Discovery Startup — SkimNews
Get the Tech newsletter
Daily tech — startups, AI labs, chips, the launches that shape the next decade. Free.
- Potential Sciences launched on November 5 with tens of millions in seed funding from SovAI and Twin Path Ventures to build AI models for drug discovery using physics-based simulations
- Potential Sciences plans to simulate atomic and molecular interactions to create large-scale synthetic datasets, aiming to overcome the limitations of small experimental databases like the Protein Data Bank
- Hatem Helal, CEO of Potential Sciences, says the goal is to make ineffective drug candidates fail faster and enable bespoke, biology-specific treatments for rare and complex diseases
- SovAI, the UK government’s AI venture fund, invested in Potential Sciences and will provide access to the national AI supercomputer, giving the startup rare computational scale for a seed-stage company
- Peter Wolynes, a chemistry professor at Rice University, acknowledges AI will improve drug discovery but compares current protein engineering to 'Roman' rather than 'Renaissance' level, underscoring the technical gap
- Alan Fersht, an emeritus professor at the University of Cambridge, supports the simulation approach, saying accurate prediction of molecular binding could revolutionize drug development if achieved
Why it matters: Pharmaceutical R&D could become faster and cheaper if AI models trained on synthetic physics-based data reliably predict drug efficacy, making rare disease treatments economically viable—though clinical testing remains a long-term bottleneck. The UK gains a strategic foothold in AI-driven science via SovAI’s $800M fund and supercomputing access.
Ask SkimNews
