【医学部学术讲座】——Big Data Sciences for Personalized and Precision Medicine
主讲嘉宾： 周小波 教授
Wake Forest University，School of Medicine
时间： 2016年12月19日上午 10:20 – 11:50
主持人： 常春起 教授
Xiaobo Zhou(周小波) 国家千人计划特聘专家 received his Bachelor degree in Mathematic at Lanzhou University 1988, and obtained Master and Ph.D. Degree at Peking University 1995 and 1998, respectively. He was an assistant professor at Harvard University in 2007, an associated professor at Cornell University from 2007 to 2011. Now he is a full professor of Diagnostic Radiology and Cell Biology at Wake Forest University。He is the directors of Center for Bioinformatics & Systems Biology, Biomedical Informatics Program, Wake CTSA and principle investigator of 5 labs in USA. Prof. Zhou’s research focuses on bioinformatics and, specifically, computational systems biology and bioimaging informatics. He has made significant contributions in gene microarray analysis, as well as bioimaging analysis and data modeling in high-content screening for compound screening, drug target validation, and candidate biomarker selection. He is currently co-investigator on two NIH funded grants.
PMI is making it increasingly feasible for physicians to prescribe the right drug, at the right dose, at the right time according to the makeup of their patient’s genome, making genome informed clinical decision support technologies as a reality. In this talk, the personalized and precision medicine by integrating huge genome and EMR will be presented.
The explosive growth of biomedical Big Data provides enormous opportunities to revolutionize current clinical practices, personalized and precision medicine if the accompanying challenges of genome, ontology, heterogeneity in those data can be addressed with novel informatics technologies. To address these challenges, we are developing the Big Data for Personalized Medicine (BD4PM) system, a biomedical Big Data informatics platform that allows fast adaptation of EMR, genome and other novel datasets by systematic data harmonization and knowledge management mechanisms, and expedites personalized and precision medicine by providing user-oriented toolkits established on a scalable translational knowledge library.
Our working hypothesis is that the similarity of genome and EMR signatures shared by patients and cell lines can reveal the underlying mechanisms of drug responses, and thus can be used in integrated models to optimize personalized medicines. We seek to investigate and develop a complete set of tools and resources that will enable signatures to be most effectively extracted from big data and applied to personalized medicine. Specifically, we unite information fusion and knowledge maturation strategies to address the challenges of data heterogeneity so that large scale signature extraction becomes feasible. We then develop novel algorithms to extract structured signatures from the integrated data to enable knowledge discovery. Finally, we investigate signature-based approaches to integrate advanced multi-scale models for personalized medicine. Through these innovative strategies, we will provide the biomedical science community with a complete set of tools and resources that allow researchers to develop and grow biomedical signatures as critically important knowledge for significant discoveries and in the process bridge key gaps between the growing amount of biomedical Big Data and the needs of translational research.