The Juvenile Diabetes Research Foundation (JDRF), a global funding agency for research on type 1 diabetes (T1D), has awarded ISE Associate Professor Shuai Huang a research grant to advance the accuracy and equity of genetic risk prediction of type 1 diabetes.

The grant supplies his team with $300K over a two-year time frame to  develop T1D Genetic Risk Score (GRS) Models for multi-ethnic populations.  Shuai’s team will develop an open-source repository that other diabetes researchers can access for analysis and creation of accurate and equitable prediction models. “Risk prediction of T1D has advanced greatly in the past decades, thanks to the rapid advancement of genotyping technologies, but recent studies have shown the state-of-the-art models have uneven prediction performances across different ethnic groups. Past research has heavily relied on data collected from research participants of European descent; an artificial intelligence (AI) and machine learning pipeline needs to be created that can recognize this dataset imbalance and properly exploit the commonality of different groups while preserving group-specific patterns. The goal is to create data equity.” 

 When asked what drew him to this research Shuai said simply, “Teddy (Bears),” but he further explained that his academic career started at the University of South Florida where he started to work with TEDDY: The Environmental Determinants of Diabetes in the Young (TEDDY) Study.” The TEDDY Study is one of the biggest international efforts to collect big data for T1D research. This study tracks more than 8000 kids from birth to 18-years-old, collecting a wide range of data from genetics to nutrition to psychological metrics. When asked why he was and is interested in T1D, Shuai explained, “as a data scientist, I simply can’t resist the mystery of such big data. Type 1 diabetes is a disease that is like a chain reaction with causes and intermediate checkpoints and outcomes.” For Shuai, it is exciting to imagine what can be achieved if he and his team can elucidate all the pathways and chain reactions; the data is just waiting for the right questions to be asked.

While the work is exciting, the collaborations are even more so. “Think about the opportunity of getting to work with people that want to solve this problem and how each is an expert in their field,” Shuai said. “It is so energizing to work with such great people. I was very lucky to meet many great scientists and medical doctors who mentored me along the way to understand the data and how the research questions could be meaningfully formulated.” Shuai’s interdisciplinary research team includes a computer scientist (Prof. Xiaoning Qian from Texas A&M), an MD (Prof. Richard Oram from University of Exeter), an epidemiologist (Prof. Kendra Vehik from the University of South Florida) and himself as project lead and statistician. The team has been working together for the past ten years. This is the third research grant JDRF has funded for their effort of advancing AI and T1D big data science research since 2014. 

Award Title: A Multi-task Learning Framework for Developing Genetic Risk Score (GRS) Models of T1D for Multi-Ethnic Populations

Project Sponsor:  Juvenile Diabetes Research Foundation (JDRF)