Predictive Modeling of the EGFR-MAPK pathway for Triple Negative Breast Cancer Patients
Overview
-
Project Name
- Predictive Modeling of the EGFR-MAPK pathway for Triple Negative Breast Cancer Patients
-
Project Center
- University of North Carolina at Chapel Hill
Project Description
The heterogeneity of breast cancer has hindered the development of predictive pathway based computational models because most approaches do not account for this disease heterogeneity, and/or use experimental data from a single cell line or animal model that is then extrapolated to most breast cancer patients. Our approach is to acknowledge and account for disease subtypes using a diverse experimental model system, and then approach the computational model building using two distinct and complementary methods to build models for a specific key tumor subtype (i.e. Triple Negative Breast Cancer (TNBC)). To accomplish these goals this project brings together a diverse scientific team, innovative computational and experimental techniques, and a unique animal model system. The project also builds on our previous findings that include a molecular taxonomy of breast cancers, which divides tumors into at least 4 disease subtypes (Basal-like, Claudin-low, HER2-enriched, and Luminal). Of note, the Basal-like and Claudin-low subtypes are most commonly classified as TNBC, which is a clinical disease subtype that lacks molecularly targeted therapies, but for which activation of the EGFR-MAPK pathway appears common. Our technological advances in proteomics have allowed us to capture and quantitate a near complete protein kinome profile on individual samples, including dynamic measurement of kinases in the MAPK pathway. Finally, we have adopted a unique TP53-/- Genetically Engineered Mouse Model system that has provided us a resource for the detailed evaluation of multiple breast cancer subtypes using a genetically controlled in vivo model system, represented by 10s of individually arisen tumors with stable and diverse phenotypes. We propose to leverage these resources, and data from two human clinical trials, to build predictive models of the EGFR-MAPK signaling pathway for TNBC patients. We will simultaneously use mechanistic and statistical modeling approaches, at a variety of scales, that incorporate data from drug treated tumors and cell lines, assayed for gene expression, DNA copy number, DNA mutations, and protein kinome activity. Each modeling approach will inform the other, thus allowing creation of two complementary multi-scale models. Lastly, we will test these computational models on human breast tumors treated with EGFR or MEK inhibitors to evaluate their relevance for human disease.
Investigators
Timothy Elston, PhD (co-PI)
Dr. Elston is a Jeffrey Houpt Distinguished Investigator in the Department of Pharmacology at the University of North Carolina at Chapel Hill and co-Director of the UNC Computational Medicine Program. His current research integrates computational approaches, including mathematical modeling and quantitative image analysis, with experimental investigations to understand complex cellular behavior. In particular, his research focuses on understanding the molecular mechanisms that regulate cell signaling networks and how these networks become dysregulated in human diseases, such as cancer.
Chuck Perou, PhD (co-PI)
Dr. Perou is the May Goldman Shaw Distinguished Professor of Molecular Oncology in the Department of Genetics at the University of North Carolina at Chapel Hill and co-Director of the UNC Computational Medicine Program. His research crosses the disciplines of genomics, genetics, cancer biology, bioinformatics, epidemiology, and clinical research. A major contribution of his has been in the characterization of the genomic diversity of breast tumors, which resulted in the discovery of the intrinsic subtypes of breast cancer. In particular his research is focused on using genomic and genetic data to build prognostic and predictive models for cancer patients.
Gary L Johnson, PhD (co-Investigator)
Dr. Johnson is a Kenan Distinguished Professor in the Department of Pharmacology where he served as chair for 14 years from 2003-2017. He is co-director of the Program in Molecular Therapeutics for the Lineberger Comprehensive Cancer Center and director of the Human Genome RNAi/CRISPR Screening Facility. His research focuses on understanding the behavior of the kinome en masse in cancer. His laboratory has developed chemical proteomics methods that allows measurement of the functional state of ~90% of the kinome that can be applied to cell lines, preclinical animal models, patient-derived xenografts and clinical trials. He integrates kinome proteomics with next generation sequencing and chromatin epigenetics to define the dynamic behavior of the kinome at both baseline and with perturbation in preclinical cancer models and patient trials.
Steve Marron, PhD (co-Investigator)
Dr. Marron is the Amos Hawley Distinguished Professor of Statistics and Operations Research, Professor of Biostatistics and Adjunct Professor of Computer Science at the University of North Carolina, Chapel Hill. His research lies in many areas statistics, data science and machine learning, with a special emphasis on gaining simultaneous insights from very diverse data types, including genomics, genetics, imaging and demographics. He enjoys using deep concepts from diverse mathematical areas including geometry and topology in novel data analyses.
Shawn Gomez, EngScD (co-Investigator)
Dr. Gomez is a Professor in the Joint Department of Biomedical Engineering at UNC-Chapel Hill and North Carolina State University and the Department of Pharmacology at UNC-Chapel Hill. He is the Director of FastTraCS, a component of the NC Translational and Clinical Sciences Institute, focused on accelerating innovation in medical devices, diagnostics, and therapeutic translational research at UNC and the broader NC healthcare system. His research focuses on integrating computational and experimental approaches towards improving our understanding of the architecture and dynamics of cell signaling networks, with particular application to cancer and infectious disease.
Andrew Nobel, PhD (co-Investigator)
Andrew Nobel is The Robert Paul Ziff Professor of Statistics and Operations Research
at the University of North Carolina. He is a senior member of the Lineberger Comprehensive Cancer Center, and a founding member of the UNC Computational Medicine Program. His research brings ideas and expertise from network analysis and machine learning to bear on problems of statistical genomics that arise in both research and clinical settings. Recent work includes the identification of differential network structures in high dimensional genomic data, and network-based approaches to expression quantitative trait loci (eQTL) analysis.