Abstract
Statistical learning models have been utilized in analysis of omics based molecular signatures predictive of disease risk and patient outcomes. Harnessing the power of big data and statistical science has brought huge benefit to patients, although the vision of precision medicine and data science is far from realized. I will introduce statistical developments and then three real studies published in the last decade. First is on a 10-year study on patient survival outcome of advanced head and neck cancer after chemotherapy evolving from high dimensional genomic data to a panel of biomarkers predictive of survival. The second case study is to assess clinical outcomes of multiple myeloma patients receiving immunotherapy where none of the biomarkers are predictive of progression survival but it is how the immune-response change over time that is predictive of patient outcome utilizing a sophisticated functional model. The last one is to predict the risk of developing Alzheimer early using metabolomics analysis. Statistical learning algorithms have played an important role in these studies. To realize the potential of current experimental and real world data, greater collaboration among researchers from multiple disciplines is necessary. (This work is a collaboration of biologists, physicians and biostatisticians and bioinformaticians)
About the Speaker
Dr. Tan is Professor and Chair of the Department of Biostatistics, Bioinformatics and Biomathematics at Georgetown University Medical Center and directs Biostatistics at its Lombardi Comprehensive Cancer Center. Previously he was Professor and Head of Division of Biostatistics and Bioinformatics at University of Maryland School of Medicine and its Greenebaum Cancer Center, (2002-2012), He was previously on faculty at St. Jude Children's Research Hospital Cancer Center and (1997-2002), and The Cleveland Clinic (1990-1997). He received his Ph.D. in Statistics in 1990 from Purdue University, Indiana. Dr. Tan’s research covers the design and analysis of clinical trials, drug combinations, biomarker evaluation, genomics and epidemiological research as well as Bayesian and semi-parametric statistical methods. Dr. Tan has served on multiple national panels including being a member of FDA Advisory Committee. He is a Fellow of the American Statistical Association and Executive Editor of Molecular Carcinogenesis, and Associate Editor of Statistics in Medicine. He has more than 200 peer reviewed publications split about evenly between statistical and biomedical journals.
Poster
