Digital Twins and the Discovery of New Drugs for Rare Diseases

The emergence of combined multi-omic and clinical data directly from patient tissue, along with the availability of ultra-scale computing, and recent developments in AI that can reverse-engineer causal mechanisms from observational data, have converged to enable the creation of Gemini Digital Twins, replicas of human disease biology. Gemini Digital Twins are computational representations of human disease that capture a critical mass of the known and unknown genetic and molecular interactions that causally drive clinical and physiological outcomes. “Virtual experiments” or simulations are conducted on these human disease replicas to discover hidden drivers of disease progression and drug response. This approach is especially important in diseases such as neurodegenerative diseases and rare diseases where the preclinical models–animals, cell lines, stem cells–are especially poor proxies of human disease. When this technology and approach is applied to rare diseases, it has the potential to discover and develop new therapeutics that can effectively treat disease in ways that are significantly more effective than has been possible to date. While rare diseases such as sickle cell anemia and Huntington’s Disease have known disease drivers at the genetic level, discovering and developing effective drugs has been elusive as “the cause is not always the cure. “ The Huntington gene was discovered by Jim Gusella at Massachusetts General Hospital over 30 years ago as the definitive genetic cause of the disease, but we still do not have an effective disease-modifying therapy. Here we will present an approach and early evidence of the application of this approach to Huntington’s Disease that has yielded a compelling discovery program that is advancing towards the clinic. The approach may be the first true example of a drug candidate moving into clinical development based on an entirely hypothesis-free approach to unraveling previously unknown human biology.