David Baker’s lab at the University of Washington has unveiled AI-driven approaches to design binders targeting intrinsically disordered regions (IDRs) of proteins, overcoming long-standing challenges in the druggability of flexible proteins. Published in Science, their amino acid sequence-based methods illustrate applications to proteins involved in pain signaling, cancer, and diabetes. Baker’s RFdiffusion model generates novel proteins to bind these regions without requiring structural data. This breakthrough expands the therapeutic target landscape to previously untargetable ‘undruggable’ proteins and opens new avenues for precision drug design.