Google releases new AI models for drug discovery

Google gave an update on how it is using AI to advance science and build products at its recent Check Up event.

In a blog post, Google’s chief health officer, Dr. Karen DeSalvo, said that since the company launched AI Overviews last year, people are more satisfied with their search results and are asking longer, more complex questions. 

AI Overviews appear in Google Search results to allow people to find general information about their inquiry from a range of sources and discover relevant sites on the web.

“And with recent health-focused advancements on Gemini models, we continue to further improve AI Overviews on health topics so they’re more relevant, comprehensive and continue to meet a high bar for clinical factuality,” DeSalvo said.

Google also announced the launch of its new medical records APIs globally in Health Connect. 

The APIs allow apps to read and write medical record information like allergies, medications, immunizations and lab results in standard Fast Healthcare Interoperability Resources (FHIR) format. 

With these additions, Health Connect supports more than 50 data types across activity, sleep, nutrition, vitals and medical records.

Last month, DeSalvo highlighted how the company received FDA clearance for its Loss Of Pulse Detection feature available on Pixel Watch 3. 

The feature can detect when there is a loss of pulse due to cardiac arrest, respiratory or circulatory failure, overdose or poisoning. It automatically prompts a call to emergency services.

The feature is available in 14 countries and will be rolled out in the U.S. at the end of March.

Additionally, with the aim of helping biomedical researchers create novel hypotheses and research plans, Google launched an AI co-scientist, a new system built on Gemini 2.0. 

The aim of the AI co-scientist is to help researchers parse large volumes of scientific literature and generate high-quality, novel hypotheses. 

“For instance, let’s say researchers want to better understand the spread of a disease-causing microbe. They can specify this research goal using natural language, and the AI co-scientist will propose testable hypotheses, including a summary of relevant published literature and a possible experimental approach,” DeSalvo said.

The collaborative tool is also aimed at helping experts uncover new ideas and accelerate their work. 

“We’re already working with partners, including Imperial College London, Houston Methodist and Stanford University, and are keen to see how researchers around the world use this tool. And while it’s in its early days, the enthusiasm is clear, we’ve received considerable interest for our upcoming trusted tester program,” DeSalvo said.

Google also announced TxGemma, a collection of Gemma-based open models intended to improve the efficiency of AI-powered drug discovery. 

“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,” DeSalvo said. 

She pointed out that TxGemma has the ability to understand regular text and the structures of different therapeutic entities that include small molecules, chemicals and proteins. 

“This means researchers can ask TxGemma questions to help predict important properties of potential new therapies, like how safe or effective they might be,” DeSalvo said. 

Finally, with Google’s assistance, the Princess Máxima Center for pediatric oncology in the Netherlands is developing an AI tool called Capricorn that utilizes Gemini models to help physicians accelerate the identification of personalized cancer treatments by combining vast public medical data and de-identified patient data.

According to DeSalvo, Capricorn quickly generates summaries of treatment options and relevant medical publications that allow physicians to have more thorough discussions on how to attain the best health outcomes for their pediatric patients. 

THE LARGER TREND

In February, Google announced a partnership with the Institute of Women’s Cancers, founded by France’s cancer research and treatment center, Institut Curie, to study how AI tools can help address cancer, share science-based health information and support postdoctoral researchers with funding. 

The partnership, which was announced at the France AI Action Summit, combines Institut Curie’s proficiency in research and medicine with Google’s advanced AI capabilities.

Primary care booking platform Healthengine also added a new booking pathway in partnership with Google. The Reserve with Google feature permits patients to book appointments directly from a practice’s Google Business profile upon search.