Date of Award
5-2024
Document Type
Dissertation
Degree Name
Doctor of Philosophy (PhD)
Department
Chemical Engineering
Committee Chair/Advisor
Marc Birtwistle
Committee Member
Jessica Larsen
Committee Member
Adam Melvin
Committee Member
Jon Calhoun
Abstract
Cancer is one of the leading causes of disease related death worldwide. Since the discovery of the genomic origins of cancer, targeted therapy has been developed towards specific mutations implicated for oncogenic transformation. However, current standard-of-care for mapping cancer patients to efficacious drug combination is often inadequate. The pathophysiology of tumor progression relies on the dysregulation of biomolecular pathways of which the topology and the dynamics challenge prognosis. Moreover, the overall genomic instability involved in disease states and the resulting inter-patient as well as intra-tumoral heterogeneity challenge rationalization of therapy and clinical decision-making. It highlights the need for the use of quantitative methodologies that may forecast clinical outcomes considering the complex nature of disease progression. In this work, we evaluated the use of single cell mechanistic modeling in predicting anticancer drug response. We begin our work with the foundation of one of the largest single cell models of stochastic proliferation and death signaling. It incorporates several signal transduction pathways which are implicated in oncogenic transformation and describes how the coordinated dynamics of these pathways drives stochastic outcomes of cellular processes such as proliferation or death in response to growth stimulus and drug dose. We addressed several aspects which may contribute to its future development towards a framework for generating unbiased drug response prediction with a more inclusive biological context encompassing multiple tumor types. At first, we focus on enhancing the accessibility and computational efficiency of the model by introducing a scalable and modular format for its construction and potential expansion. Then, we developed a mechanistic cell population simulation framework based on the single cell simulation functionality of the model. This allowed us to generate representations of dynamic cell populations, bridging the gap between simulation outputs and experimental datasets, such as dose response for various drugs. A direct comparison of simulation outputs with experimental datasets enabled validation of the current modeled biology as well as identification of crucial knowledge gaps within the ERK signaling and cell cycle pathways. Furthermore, we developed a method to perform omics-informed context definition taking inputs of genomic, transcriptomic, and proteomic datasets for a number of cancer cell lines from one of the largest datasets of cancer cell characteristics, the Cancer Cell Line Encyclopedia. This allowed us to generate cell-line specific model variants as well as devise a strategy for the mechanistic exploration of drug sensitivity datasets generated for these cell lines. We believe the methods presented here will help provide guidance in attempting to build a deeper quantitative understanding of the dynamic and multivariate molecular complexities that currently challenge treatment efficacies in cancer.
Recommended Citation
Mutsuddy, Arnab, "Single Cell Pharmacodynamic Modeling of Cancer Cell Lines" (2024). All Dissertations. 3572.
https://tigerprints.clemson.edu/all_dissertations/3572
Author ORCID Identifier
0000-0001-6488-7067