The two organizations join forces to develop pioneering AI-based solutions for efficient chemical synthesis planning that accelerates the development of new drugs.
CAS, a division of the American Chemical Society specializing in scientific information solutions, and Molecule.one, a technology leader in AI-based solutions for pharmaceutical chemistry, have established a strategic collaboration focused on the joint development of computer-aided design synthesis technologies to accelerate scientific advances in early-stage drug discovery and assist chemists in the discovery of new small molecules.
Leveraging its existing technologies and expertise, including Molecule.one’s generative deep learning models and synthesis planning platform with a chemist-first user interface, and world-class collection of chemical reaction content and the extensive knowledge of the CAS industry, the two organizations will work together to improve and develop solutions for efficient and innovative chemical synthesis planning. Beyond the complementarity of their capabilities, the collaboration between the two organizations is driven by their common goal of empowering scientists and accelerating progress.
“We are delighted to bring to market the first generative models for chemistry trained on CAS content, as a strategic partner of CAS for synthetic accessibility and retrosynthesis deep learning enhancements,” said Piotr Byrski, Co-Founder and CEO of Molecule.one. . “Generative AI has shown amazing feats in multiple fields when trained on large data sets and I think there is a clear need to bring this to drug discovery. I am very proud of the CAS and Molecule.one teams, who have worked diligently over the last year to identify how we can bring together Molecule.one’s unique technology and expertise in AI and CAS’s extensive chemical data set.
Also Read: KBI BioPharma Continues to Strengthen Leadership and Expertise with Key Executive Appointments
The first joint offering, M1 RetroScore powered by CAS, will be part of the Molecule.one suite of products and will replace its current synthetic accessibility scoring tool, RetroSAS. Leveraging recent advances in generative model development by the Molecule.one team, the solution uses machine learning models trained on the content of the best CAS chemical reactions to predict the probability of synthesis of new molecules. M1 RetroScore powered by CAS provides all users with synthetic accessibility scores along with corresponding potential synthesis pathways for given targets. For CAS SciFinder n users , the tool will connect with reference reactions within the CAS SciFinder platform.no . M1 RetroScore powered by CAS is the first commercially available synthetic accessibility scoring tool based on CAS content.
Molecule.one and CAS are also collaborating to enhance the world-class retrosynthesis capabilities already available in CAS SciFinder n . As part of its continued dedication to improving its solutions, CAS will incorporate Molecule.one’s deep learning models to improve its synthesis predictions and help scientists explore potential synthetic pathways for small molecules, thereby accelerating scientific breakthroughs.
“The chemical reactions and retrosynthesis capabilities of CAS are unmatched,” says Tim Wahlberg, CAS product manager. “By collaborating with Molecule.one, we will bring to market a unique combination of capabilities to solve the challenges posed by synthesis planning, both in the initial phase of searching for synthesisability and in the design of synthesis in the laboratory and at scale. These solutions will continue to reduce the time scientists spend in the discovery process and allow them to be more efficient throughout the entire innovation process.”
SOURCE: BusinessWire