Data Science and the Fight Against Cancer - News from the Frontier

Seminar Series:      Bioinformatics Seminar    

Date:      Tuesday, April 2, 2019

Time:      11:30 a.m. – 12:20 p.m.

Location:      BRNG 2290

Speaker:      Eric Stahlberg, PhD

Affiliation/Organization:      Frederick National Laboratory for Cancer Research, Leidos Biomedical Research

Abstract: The convergence of new technologies to probe biology, methods to amass biomedical information, and scalable approaches to analyze the growing volumes of data are shaping the frontiers for cancer research. The importance of data science continues to grow as these frontiers are reached, with new challenges encountered and opportunities identified. Deep learning is one of the methods that has experienced a rapid growth in the development of new analytic workflows and creation of predictive models. Deep learning is extending other machine learning approaches and opening the door to new approaches to make predictions with large scale data in many areas including tumor response to drug treatments. The talk will share some of the latest developments in the NCI-DOE collaboration for the Joint Design of Advanced Computing Solutions for Cancer, with special attention to CANDLE (the CANcer Distributed Learning Environment) and its early application to creating predictive drug tumor response models. Challenges encountered and opportunities envisioned in the development, sharing, evaluation and use of predictive models will also be discussed.  

For more information and associated reading, please see: http://www.stat.purdue.edu/~minzhang/598_Spring2019/schedule.htm.