Google TxGemma, an open LLM for drug discovery, in late March | Large language models ai | Llm chatbot | Large language models course free | Turtles AI

Google TxGemma, an open LLM for drug discovery, in late March
Google Announces New AI Models to Support Biomedical Research and Health Data Management
Isabella V18 March 2025

 

The Mountain View company introduces TxGemma, a set of open AI models designed to accelerate pharmaceutical research by improving understanding of the properties of therapeutic molecules. Updates to Health Connect and an AI co-scientist launched to support biomedical research. New APIs for medical records and advanced capabilities in digital health research expand the landscape of technological innovation in the medical sector.

Key points:

  • TxGemma and drug discovery: new AI models to predict the properties of therapeutic molecules.
  • AI co-scientist: support research with literature analysis and hypothesis generation.
  • Health Connect and medical records: global APIs for standardized health data management.
  • Pixel Watch 3: advanced loss of pulse detection with FDA approval.

Google announced the launch of TxGemma, a collection of open AI models for drug discovery, with the aim of improving the analysis of the properties of chemicals, molecules and proteins. Integrated into the Health AI Developer Foundations platform, these models allow researchers to ask questions to obtain predictions about the safety and effectiveness of new therapies. The initiative is part of a broader development of AI tools for health, including an innovative co-scientist based on Gemini 2.0. The latter is designed to support biomedical experts in the analysis of large quantities of scientific publications, facilitating the formulation of new hypotheses and experimental plans.

In the context of healthcare digitalization, Google has expanded the support of medical records in its Health Connect service, introducing APIs that allow the structured management of medical information such as allergies, medications and vaccinations in the standard FHIR format. This update, available globally, enriches an ecosystem already capable of processing data relating to vital signs, physical activity and nutrition. The company has also enhanced the AI ​​Overviews function in Search, improving the quality of health information accessible to the public and expanding its availability to new languages ​​and countries.

Another new feature is the Pixel Watch 3, which will introduce the Loss of Pulse Detection feature in the United States. This system, authorized by the FDA, automatically detects the absence of a heartbeat and initiates an emergency call to ensure timely intervention.

With these innovations, Google strengthens its commitment to the healthcare sector, providing advanced tools that improve research and management of clinical data.