Computational Approaches to Delineate Complex Genetic Variants of Cancer Genomes for Precision Oncology
ABSTRACT: Tens of thousands of molecular profiles of cancer patients and preclinical models have been generated and analyzed for cancer research. These “Big Data” analyses have provided the unprecedented opportunity to translate knowledge into the discovery of novel therapeutic targets and treatment strategies for cancer patients. Although significant efforts have focused on genetic variations that disrupt the coding sequence of important cancer-related genes, the role of complex genomic variation and rearrangements that present in either coding or non-coding is still largely unknown for therapy response in cancer. In this talk, I will present our ongoing efforts to develop novel computational tool and machine learning algorithms to (1) predict individual tumor sensitivity to drug(s) and chemical compound(s) based on information on genetic variants; (2) stratify patients into clinically distinct subgroups and predict risk of survival based on an integrative large scale somatic mutation analysis. Our results show that incorporating information from complex genetic variants can identify a predictive biomarker-based on genetic variations that can guide therapy for patients with cancer in the clinical setting. Moreover, our results demonstrated that the predictive model that integrates genetic variations of clinically relevant cancer-related genes and multiple drug sensitivity profiles could improve prediction of the sensitivity of a panel of FDA-approved anticancer drugs and novel synthetic compounds. Finally, our results also showed that a large-scale integrative somatic mutation analysis could improve to better stratify patients into clinically distinct phenotypes and accurately predict the risk of survival after surgery. Our computational algorithms and findings could improve our understanding of how complex genetic variations affect clinical decision-making and treatment stratification, and thus improve therapeutic efficacy in cancer.
BIO: Tae Hyun Hwang is a faculty member at Cleveland Clinic where he currently leads and oversees bioinformatics and computational science research. His research interest is developing new machine learning and data mining algorithms that might be used clinically to guide therapy decisions for patients. He leads multi-institute/center bioinformatics research to study human cancer and serves as a bioinformatics core director for A NASA Specialized Center of Research (NSCOR) and data analysis core co-director for the University of Texas Kidney Cancer the Specialized Programs of Research Excellence (SPORE) grant. He is a recipient of Lung SPORE Career Development Award and American Cancer Society Young Investigator Award.