The Michigan Institute for Data Science received a $2.3 million grant from the National Institutes of Health this month to develop a national training program designed to equip faculty and scientists across the United States with the skills necessary to enhance the rigor and reproducibility of biomedical research.
Rigor and reproducibility are crucial, yet challenging, components of scientific research, particularly in fields involving complex data types from varied sources and intricate data manipulation pipelines, such as biomedical data science.
Whether research data, analytical methods and research workflows are rigorously designed and implemented, and whether they are thoroughly documented and shared to enable anyone to reproduce and verify their research, is essential to the trustworthiness of science.
“Researchers are acutely aware of its importance, but frequently lack the necessary resources and technical knowledge to consistently reach optimal standards,” said Jing Liu, the project’s principal investigator and executive director of MIDAS, a unit based in the Office of the Vice President for Research.
“Recognizing this gap, MIDAS has been working diligently with data science and artificial intelligence researchers from across U-M to understand their challenges, curate effective methods, and promote improvements in research rigor and reproducibility.”
The new training program will convene a series of annual bootcamps covering key considerations in biomedical research, such as ethical issues, data management, statistical design, predictive models, reproducible workflow and meta-analysis.
The first training bootcamp, planned for summer 2024, will accommodate 100 attendees, with 50 full scholarships earmarked for trainees from minority-serving institutions, underrepresented demographic groups and institutions with limited resources.
Following these bootcamps, trainees will apply the knowledge and skills acquired in their respective research fields and develop training programs at their home institutions, with mentoring from the instructors over a year’s time.
As opposed to training individual researchers, the program will encourage teams to embrace each member’s scientific expertise and technical skills, fostering a collaborative approach to rigor and reproducibility.
“Our ultimate goal is to effect institutional transformation by enabling and sharing important training materials with groups across U-M and beyond to achieve the broadest impact,” said H.V. Jagadish, the Edgar F. Codd Distinguished University Professor of Electrical Engineering and Computer Science, the Bernard A. Galler Collegiate Professor of Electrical Engineering and Computer Science, and director of MIDAS.
Faculty and staff representing multiple U-M units will partner with colleagues from the College of William and Mary, as well as minority-serving institutions Jackson State University and the University of Texas at San Antonio, to develop and implement the program.
Reflecting MIDAS’ commitment to diversity, equity and inclusion, its team aims to accommodate trainees of different backgrounds, helping to advance responsible data science and AI research nationwide.
“With the rapid emergence of new research methods, such as generative AI, programs like this play a crucial role in ensuring ongoing scientific trust,” Liu said.
“As we adapt to this rapidly evolving research landscape, MIDAS and similar research institutes can effectively guide researchers and academic institutions to increase their competitiveness, advancing innovations and discoveries that positively impact the world around us.”