$2.5M NIH Grant Supports AI and Data Science in Diabetes Research

For years, Computer Science Professor Chen Li has been working to develop Texera, an open source platform that uses browser-based workflows for collaborative data science and AI/machine learning. Now, thanks to a new five-year, $10M multi-institutional grant from the National Institutes of Health (NIH) to UC Irvine, Cornell, UCLA and UCSD, Li will be using the Texera system to support data science and AI in the diabetes research community.

Specifically, the National Institute of Diabetes, Digestive and Kidney Diseases (NIDDK) has awarded UCI $2.5M to support a project outlining plans for the next-generation NIDDK Information Network (dkNET). The NIH NIDDK grant, “dkNET Coordinating Unit: Harnessing the Power of AI and Data Science for Collaborative Discovery and Sharing in the DK Community,” supports the centralized dkNET resource, which connects the NIDDK research community to a growing number of biomedical resources (including organisms, reagents, materials, and protocols), data and bioinformatics tools.

“It’s our first time using a home-grown open-source system to provide cloud services to community users to do data science,” says Li, “and it’s full of challenges and excitement.”

Li will be working with others to extend dkNET services, bringing powerful new AI and machine learning (ML) techniques, along with cloud computing resources, to the NIDDK research community. This will help the community better leverage data assets cataloged by dkNET, as researchers explore and develop hypotheses.

The objective is to create a unified ML paradigm for performing a diverse range of analytical tasks, such as

• processing various data types cataloged by dkNET,
• integrating data of multiple modalities from different sources, and
• incorporating domain-specific knowledge from public knowledge bases.

The resulting knowledge models from this paradigm will help researchers locate supporting or contradictory data to evaluate hypotheses.

— Shani Murray