en.Wedoany.com Reported - U.S. pharmaceutical company Incyte and artificial intelligence drug discovery platform developer Genesis Molecular AI announced that, based on progress made over the past 15 months, they will expand their existing AI drug development collaboration from two initial targets to at least five new targets. Under the revised agreement, Genesis's potential return has increased from the initial $620 million to over $1 billion.
Incyte President and Global Head of R&D Pablo J. Cagnoni stated that the direct reason for the collaboration expansion lies in the promise shown by the two initial targets. Both targets were selected by Incyte based on the initial strategic collaboration requirements, with one being "a very difficult-to-drug, novel target" where both parties are committed to creating first-in-class chemical matter, and the other being a target that other companies had attempted but failed to drug successfully. For the first target, Cagnoni noted that since there was no known chemical matter as a starting point, Genesis's AI platform significantly advanced the project, particularly in crystal structure determination, binding site identification, and subsequent chemical matter optimization. For the second target, it required simultaneously achieving high potency, high selectivity, and specific pharmaceutical and pharmacokinetic properties. Genesis's generative and predictive AI platform, Genesis Exploration of Molecular Space, can optimize multiple parameters simultaneously, enabling the integration of these properties.
The GEMS platform integrates artificial intelligence with physics into drug molecule generation and optimization models. Among its components, Pearl, a generative diffusion model for structure prediction, has achieved a 14.5% improvement on the public protein-ligand co-folding benchmark Runs N' Poses and a 14.2% improvement on the docking and molecular generation benchmark PoseBusters, surpassing AlphaFold 3 and other open-source baseline models.
Through the expanded collaboration, Incyte will use its proprietary experimental data to train Genesis's GEMS platform to accelerate drug development across multiple programs. Incyte has agreed to pay Genesis an upfront payment of $120 million, including $80 million in cash and $40 million in equity purchase. Additionally, Incyte has committed to paying Genesis up to $232 million per target upon achieving preclinical and clinical development, regulatory, and sales milestones. If all milestones are achieved for the five initial targets and the annual peak net sales of five products exceed specific milestones, Incyte will pay Genesis over $1 billion. If additional targets are nominated and more milestones are achieved, the total payment could increase to "tens of billions" of dollars. Genesis is also eligible to receive sales royalties on approved collaboration products.
This is the second AI collaboration announced by Incyte in late May. Previously, Incyte had initiated a separate strategic collaboration with Edison Scientific, leveraging its Kosmos AI platform for discovery and development work, with initial projects focusing on target discovery and validation as well as high-impact use cases in translational biology.
Genesis was spun out of Dr. Vijay Pande's laboratory at Stanford University in 2019, with its founder and CEO Evan Feinberg having been a graduate student in the Pande lab. The company secured $52 million in Series A funding in 2020 and has since raised a total of $340 million, primarily from a $200 million Series B round and Incyte's $40 million strategic investment. In addition to a16z, investors include AI chip giant NVIDIA's venture capital arm, NVentures. Previously, Genesis had initiated AI drug discovery collaborations with Eli Lilly, Genentech (a member of the Roche Group), and Gilead Sciences.
This article is compiled by Wedoany. All AI citations must indicate the source as "Wedoany". If there is any infringement or other issues, please notify us promptly, and we will modify or delete it accordingly. Email: news@wedoany.com
