One of the oft-repeated positives related to the ascendence of artificial intelligence is the idea that AI will somehow help to cure disease… specifically cancer.
While that seems like a lofty goal, some hints as to how this might occur are beginning to emerge through a series of partnerships between pharmaceutical firms and AI companies.
The latest of these partnerships was recently announced when Anglo-Swedish pharma giant AstraZeneca partnered with AI startup Absci to develop new cancer drugs.
The deal worth an estimated $250 million will focus on utilizing artificial intelligence to help predict useful antibodies for scientists to focus on that are most likely to be effective in treating specific types of cancer.
Puja Sapra, AstraZeneca’s head of research and development said of the deal, “We are applying AI throughout our discovery and development process, through building in-house capabilities and through collaborations such as with Absci.”
Absci is based in Vancouver, Washington, and was founded in 2011.
The company’s product is called Drug Creation and develops billions of protein-to-protein interactions which are used by machine learning to train the underlying generative AI model.
Machine learning has been increasingly used to help with risk assessments, early diagnosis, patient prognosis estimation, and treatment selection for many types of cancer.
Absci’s approach layers the machine learning approach to large sets of protein-related data and applies generative AI to serve to answer questions that can help scientists develop treatments more quickly and safely.
Earlier this year Microsoft announced its research subsidiary was pairing its supercomputing power with the capabilities of Paige.ai, a startup created out of Memorial Sloan Kettering Cancer Center.
That partnership is seeking to build one of the world’s largest image-based AI models for digital pathology and oncology.