Taking a potentially significant step in the use of artificial intelligence (AI) to dramatically accelerate early R&D in drug discovery, Alphabet Inc.’s Google on Tuesday said it’s developing “open” AI models to do just that.
The collection of models, called TxGemma, were disclosed at a health care event in New York.
“The development of therapeutic drugs from concept to approved use is a long and expensive process, so we’re working with the wider research community to find new ways to make this development more efficient,” Karen DeSalvo, chief health officer at Google, wrote in a blog post. “[R]esearchers can ask TxGemma questions to help predict important properties of potential new therapies, like how safe or effective they might be.”
The AI models, to be made available this month through Google’s Health AI Developer Foundations program, understand “regular text” and structures of different “therapeutic entities” that include chemicals, molecules and proteins, according to Google. It remains unclear if the models’ license will allow for commercial use or customization.
Isomorphic Labs, a Google spinoff that boasts partnerships with Eli Lilly and Novartis, has aspired to fast-track drug discovery via AI and begin AI-designed drugs this year, but progress has proved to be incremental.
Isomorphic is one of more than 460 AI startups toiling on drug discovery, while investors have ladled $60 billion into the space.
In addition to drug discovery, Google on Tuesday announced improved overviews in Google Search for health queries, and medical records application programming interfaces (APIs).
Google said it is using AI and ranking systems on search to expand “knowledge panel” answers on thousands of health-related topics, and adding support for queries in Spanish, Portuguese and Japanese on mobile devices. A new mobile search feature in the U.S., “What People Suggest,” is designed to highlight content from users with shared experiences on health conditions. In one example, people seeking advice on exercises to cope with arthritis yielded collated reports from various forums across the web.
“While people come to Search to find reliable medical information from experts, they also value hearing from others who have similar experiences,” DeSalvo wrote in a blog post. “Using AI, we’re able to organize different perspectives from online discussions into easy-to-understand themes, helping you quickly grasp what people are saying.”
Google’s new APIs, meanwhile, will enable apps to “read and write medical record information like allergies, medications, immunizations, and lab results in standard FHIR (Fast Healthcare Interoperability Resources standard) format,” DeSalvo added.
