AI-Designed Proteins Target Undruggable Sites, Opening New Frontiers in Medicine

Scientists have developed a novel method called "logos" that could revolutionize drug discovery and diagnostics by targeting intrinsically disordered regions (IDRs) of proteins, which were previously considered "undruggable." Proteins, essential for various bodily functions, often malfunction in diseases like cancer and diabetes. Traditional drug discovery focuses on stable protein structures, but this approach ignores the majority of the proteome, where more than 50% of proteins have IDRs. In a paper published in Science, researchers from David Baker's lab at the University of Washington describe logos, an AI-driven technique that designs new molecules capable of binding to these flexible regions. Unlike conventional methods that struggle to target IDRs due to their lack of a fixed shape, logos creates binder molecules that form stable pockets into which the disordered protein regions can fit. This method leverages the inherent flexibility of IDRs, using it to advantage by guiding the target sequence into a favorable binding conformation. The researchers successfully designed binders for 39 unstructured protein targets, demonstrating the potential of this approach. "Our computational design pipeline enables the design of binding proteins to arbitrary disordered peptides and proteins," the researchers noted. "Although targeting disordered proteins has been a considerable challenge, we show that the disorder is an advantage, driving the target sequence into a binding-competent conformation." This breakthrough could open new avenues for developing treatments for a wide range of diseases, particularly those where IDRs play a significant role, such as cancer and neurodegenerative disorders like Alzheimer's. By making these previously inaccessible targets viable, logos could expedite the discovery of life-saving medications and advance our understanding of complex biological processes. Industry insiders and experts are hailing logos as a game-changer in the field of medicinal chemistry. The method underscores the growing importance of AI in addressing challenging problems in biotechnology and drug development. David Baker, a Nobel Prize-winning chemist and leader of the project, emphasizes that computational design is becoming an indispensable tool for biologists and pharmacologists. Companies focused on AI-driven drug discovery, such as Atomwise and Exscientia, are likely to integrate this technology into their platforms, further accelerating the pace of innovation in the pharmaceutical industry.