precisionFDA
precisionFDA is a collaborative, cloud-based platform that’s transforming the way FDA scientists and researchers exchange information with industry, scientists, and the public, making it easier to collect, analyze, and apply data to solve real world problems.
FDA research into new areas of science informs our regulatory decisions.
Science and data are at the heart of all regulatory decisions to promote public health. As scientific advancements lead to the creation of new therapeutics, foods, and cosmetics, FDA scientists sometimes face new questions about how to regulate them. We often call on experts from different disciplines in the academic and scientific fields using precisionFDA.
Our researchers and data scientists evaluate how new areas of science are affecting FDA-regulated products. The agency integrates the results of this research into our regulatory review process, enabling our reviewers to make science-based decisions about a product’s benefits and risks.
FDA harnesses cloud technology to advance collaborative regulatory science research.
In 2015, the FDA launched the high-performance computing environment precisionFDA, using the opportunities offered by cloud technology, which makes it possible to store and access data, like files and applications, on remote servers over the internet rather than on a single computer.
Now, large sets of data can be securely shared with anyone who works with FDA. With shared technology, precisionFDA continues to advance data research, analytics, and genomics. FDA scientists, industry, academia, and the global scientific community can collaborate using the same data sets and algorithms to improve public health.
See Also:
• precisionFDA.gov
• FDA’s Global Substance Registration System (GSRS)
• OpenFDA
- Removes Barriers to Collaboration: Before precisionFDA was developed, our scientists and regulators couldn’t collaborate virtually with our external partners on extremely large datasets. These datasets had to be transferred physically, which often took months, limiting our ability to perform collaborative research. Today, precisionFDA provides our staff with a virtual lab environment that connects FDA scientists and reviewers with a global network of experts in just minutes.
- Enables Information-Sharing: precisionFDA helps us improve our understanding of evolving areas such as real-world data, genetics, and AI. This platform provides access to private shared workspaces where individuals and organizations can conduct various analyses, and a shared communal space where we publish and share research results, reference materials, and tools and algorithms.
- Makes it Easier to Solve Problems Using Innovative Methods: If something doesn’t look right in the virtual lab environment, there’s a community of experts that can help identify what went wrong, making it easier to fix. precisionFDA also furthers research through our public challenges that invite scientists and members of the global data science community to participate in solving problems. For example, recent research includes developing innovative food traceability tools and creating and evaluating AI models to predict health-related results in veterans.
- The Generative Artificial Intelligence (GenAI) Challenge: Democratizing and Demystifying AI is part two of a three-part series to make AI more accessible and understandable to a wider audience. “Democratizing AI” enables many industry sectors to obtain the benefits from AI advancements, creating a broader adoption. “Demystifying AI” will help enhance transparency, address bias in AI systems, and build trust in AI technologies. The results of this Challenge will help inform regulatory science and shape the future of Generative AI in healthcare. (January 10, 2025 - February 28, 2025)
- The Digitally-Derived Endpoints for Freezing-of-Gait Detection (DEFoGD) Challenge sought artificial intelligence (AI) models to identify and predict freezing of gait (FoG) events related to Parkinson’s disease. Medical device developers, tech innovators, software developers, and academic researchers were invited to participate, to help us better understand how digital health technologies can be leveraged to monitor and collect digitally derived endpoints, such as biometrics data collected through wearable devices. (May 28, 2024 - August 6, 2024)
The Automated Machine Learning (AutoML) App-a-thon: Democratizing and Demystifying Artificial Intelligence is part one of a three-part series to make AI more accessible and understandable to a wider audience. AutoML App-a-thon invited scientists and data analytics professionals to explore AutoML algorithms and improve our understanding of the benefits and constraints of using AutoML tools with medical data. Participants used AutoML tools on provided datasets to match or improve the performance of previously developed traditional ML models. (February 26, 2024 - April 26, 2024)