Author ORCID Identifier
Semester
Summer
Date of Graduation
2024
Document Type
Dissertation
Degree Type
PhD
College
Statler College of Engineering and Mineral Resources
Department
Chemical and Biomedical Engineering
Committee Chair
David J. Klinke II
Committee Co-Chair
Lizzie Y. Santiago
Committee Member
Lizzie Y. Santiago
Committee Member
Tracy Liu
Committee Member
Gangqing Hu
Committee Member
Margaret Bennewitz
Abstract
Quantitative Systems Pharmacology (QSP) refers to a mechanistic modeling approach used for assessing potential therapeutics by linking the molecular and cellular level mechanisms of the disease and therapeutic to the system-wide dynamics and clinical endpoints relevant to the disease. Although the field of QSP is still in its relative infancy, the mathematical methodology employed to predict the kinetics of these systems are not new and can be applied to a wide variety of systems. For instance, similar to QSP’s goal of guiding the development of novel therapeutics for various disease states, there exists a need to develop novel interventions in the engineering education space that can be guided using similar methodologies.
This body of work aims to apply Bayesian statistics, a cornerstone methodology within QSP, in both cancer immunology and first-year engineering education. The first study focuses on four secreted factors identified in a previous study that potentially mediate immunosuppression and could become targets for novel immunotherapies. We tested for clinical correlates in existing human data and verified in vivo whether knocking out tumor cell production of these factors improved immune-mediated control of tumor growth. A kinetic analysis leveraging III a Markov Chain Monte Carlo (MCMC) approach quantified the various knockouts’ effect on cells’ intrinsic growth rate. Key results suggest that CCN4 is a mediator of immunosuppression in the melanoma tumor microenvironment and a potential collateral immunotherapy target.
The remaining two studies focus on improving the performance and retention of non-calculus ready first-year engineering students. As college readiness continues to decline, the proportion of students entering engineering programs with low mathematics proficiency is increasing. These students have lower retention rates than their calculus-ready peers. We created two first-year engineering courses, one college algebra-based course focused on improving critical thinking skills and one trigonometry-based course focused on improving metacognition skills and used Bayesian methods to analyze the impact of the intervention on various student success metrics. Students in both courses saw higher pass rates in their math course compared to the control group. Those participating in the college algebra-based intervention course saw improved critical thinking skills and increased cumulative GPA and retention in engineering. Improvement of skills in the trigonometry-based course was dependent on additional factors such as the difficulty level of the problem analyzed. Preliminary data indicates that combining these courses into a two-semester long intervention could provide improved benefit to the population of non-calculus ready engineering students and improve their success in engineering.
Recommended Citation
Pirkey, Anika C., "Implementation of Bayesian Statistics to Answer Diverse Engineering Questions: Applications in Cancer Immunology and Engineering Education" (2024). Graduate Theses, Dissertations, and Problem Reports. 12533.
https://researchrepository.wvu.edu/etd/12533
Embargo Reason
Publication Pending