Combining Artificial Intelligence and Human Expertise in Pursuit of Novel PolQ Inhibitors
Ageo Miccoli, Annabelle Charnock, Alexis Denis, Euan A. F. Fordyce, Cameron Franklin, Ennys Gheyouche, Ian Henderson, Christopher Housseman, Allan Michael Jordan, Maxime Laugeois, Jean-Christophe Meillon, Stuart T. Onions, Quentin Perron, Clémentine Pescheteau, Émilie Pihan, Mark Reeves, Kirsty Rooney, Hannah Rowlands, Nicole Scally, Christopher Sleigh · ACS Omega · 2026
Application of generative AI, synergistically with traditional knowledge-based approaches, has expedited the discovery of small-molecule PolQ polymerase domain inhibitors. Leveraging AI-generated ligand designs, a structurally diverse set of compounds were synthesized and evaluated, ultimately identifying a novel oxime-based scaffold as a promising starting point. Although the initial hit demonstrated only modest biochemical potency, the oxime moietyan unexplored vector in this chemical spaceoffered an opportunity for strategic substitution. Gratifyingly, using this strategy, a highly potent array of functionalized oxime PolQ inhibitors displaying promising initial ADME properties was delivered, underscoring the power of generative AI to unlock new chemical space and accelerate lead optimization.