For its DREAM 9 challenge, the University has given 31 international teams access to a unique set of patient data from hundreds of leukemia patients receiving treatment at the University of Texas MD Anderson Cancer Center in Houston. DREAM, which stands for Dialogue for Reverse Engineering Assessment and Methods, acts as a platform for crowd-sourcing ideas for big-data tools that can solve biomedical problems.
The teams were given information about patients, such as demographic data and complex disease information, but had no insights about outcomes. The primary challenge was to determine how well the algorithms teams developed could predict patient response to chemotherapy. The best-performing team developed technology that could predict patient response with close to 80 percent accuracy.
“We used DREAM as a way to get general insight into making more accurate predictive models of clinical outcomes,” Amina Qutub, PhD, a RiceUniversity bioengineer, said in a statement. “Steve (Kornblau), who runs the core banking facility for leukemia patients at MD Anderson Cancer Center, had the foresight to start gathering and banking patient biopsy samples when he was a resident over 25 years ago. The bank is a fantastic resource and a tremendous gift to the public. Genomic and proteomic analysis on a portion of these patient biopsies served as the basis for DREAM.”
Results of the challenge are published in PLOS Computational Biology. Dr. Qutub’s lab is currently using insights that came from the DREAM challenge to investigate new leukemia treatments.
More articles on health IT:
Update: Hacker now selling 9.3M records on dark web, threatens more data dumps
How big data helped uncover the largest Medicare fraud takedown
Patient control is key to data exchange, but faces cultural, economic barriers
