Cancer Systems Biology Scholar Program,
Department of Radiology/Biomedical Data
Science, Center for Cancer Systems
Stanford University
ABSTRACT
Single-cell technologies are being used to generate massive amount of data to characterize the biological heterogeneity of normal and pathological tissue. The ability to obtain highly resolved molecular phenotypes directly from individual cells is transforming the way we define cell states and study cellular responses to perturbations such as small molecules, cytokines and immunotherapies. Recent studies have shown that intratumor heterogeneity influences the prediction of effective drug combinations. The increasing number of FDA approved targeted drugs and associated resistance mechanisms due to intratumor heterogeneity warrants the need for a systematic and personalized combination therapy that will reduce the risk of drug resistance and patient relapse. Drug combination prediction methods have mainly focused on targeting the predominant tumor subpopulations. Single-cell technologies such as CyTOF reveal the heterogeneity inherent in primary tissues and provide the means to characterize complex phenotypes, isolate rare populations and precisely target particular subpopulations. A key advantage of single-cell mass cytometry, is that many surface and signaling markers can be simultaneously measured. The proposed research provides a novel computational framework called DRUGNEM that can be used within a clinical setting to integrate single-cell single-drug screening data for the purpose of identifying the optimal drug combination that maximizes the treatment response across a heterogeneous population of cells based on intratumorsignaling changes associated with a positive phenotype such as cell death measured by changes in specified death markers. DRUGNEM identifies sub populations of cells based on lineage markers and a nested drug network based on their effects on intracellular signaling markers that may be similar or different between these sub populations. Drug combinations that offer the maximal treatment response with the minimal number of drugs are ranked the highest. Using Bcr-Scr, JAK/STAT and PI3k/mTOR inhibitors, DRUGNEM identifies different optimal regimens for each of the 30 Ph+ALL pediatric patient samples. Also, from a theoretical perspective, my research provides preliminary evidence that synergy at the bulk population level is different than aggregated across single cells.
Host: May Wang, Ph.D.
Tuesday, January 24
10:30 a.m.
McIntire Room 3115,
Whitaker Bldg.
Videoconference:
Emory: HSRB E182
Georgia Tech: TEP 104