Date of Award
Fall 2023
Document Type
Dissertation
Degree Name
Doctor of Philosophy (PhD)
Department
Computational Biology and Bioinformatics
First Advisor
Townsend, Jeffrey
Abstract
The progression of cancer, including the acquisition of therapeutic resistance and the mortality-inducing metastatic spread of therapy-resistant cell populations, is fundamentally an evolutionary process. To contribute to the development of effective treatments against this evolving scourge, this work leverages phylogenetics, statistics, and machine learning techniques to quantify the effects of genetic variants responsible for therapeutic resistance, resolve the timing of their occurrence, understand evolutionary trajectories, and profile the role of neoantigens in varied cancer systems. First, a bioinformatics pipeline for quantifying selection on neoantigens—tumor-specific antigens and potential targets of the immune systems—is detailed. Second, the evolutionary history of two patients with metastatic lung cancer is dissected, revealing shifting premetastatic endogenous and exogenous mutational processes and their effect on metastatic progression and therapeutic resistance. Third, a large cohort of patients with identical qualifying mutations for a targeted therapy are used calculate the first ever strength of selection on variants in vivo, compare them to variants driving primary progression, and quantify therapeutically driven immune selection pressures—resulting in the identification of orthogonal targets for coupled immune therapy in lung adenocarcinoma. Fourth, the convergent and parallel evolution of non-small cell lung cancer metastases to the famously immune-privileged brain are similarly probed, revealing a more complex and diffuse selective pressure at work. Lastly, the limits of phylogenetic approaches for predicting ancestral states to characterize evolutionary trajectories and etiology are encountered and surmounted with supplementary application of machine learning in clear-cell renal-cell carcinoma. These essential insights illuminate the mechanistic evolution of tumorigenesis, therapeutic resistance, and metastatic progression in the selected systems while creating a powerful, general framework for understanding and predicting other cancer systems.
Recommended Citation
Fisk, J Nic, "Computational and Evolutionary Approaches for Translational Cancer Research" (2023). Yale Graduate School of Arts and Sciences Dissertations. 1179.
https://elischolar.library.yale.edu/gsas_dissertations/1179
