Novel applications of expected utility theory to epigenetics, signal detection and epidemiology
Description
TitleNovel applications of expected utility theory to epigenetics, signal detection and epidemiology
Date Created2021
Other Date2021-05 (degree)
Extent1 online resource (ix, 80 pages)
DescriptionTrade-offs occur at every level of ecological organization and are sensitive to changing environments and sudden perturbations to systems. To study trade-offs, it is necessary to characterize the costs and benefits of traits under conditions of environmental change, however this can be challenging to accomplish empirically. Economic theory has a long history of examining optimal strategies in games, gambles and investments. An economic framework that characterizes costs and benefits of a particular strategy is called expected utility theory. Within the body of this work, we propose that expected utility theory may be applied to anticipate how an individual or population will respond to changing environmental conditions. We apply this expected utility theory framework to three biological systems to explore optimal decision making in the face of trade-offs to individuals and populations in the face of changing environmental conditions. Specifically, we model expected utility of epigenetic gene regulation, expected utility of signal detection criteria in social populations, and expected utility of social isolation during a pandemic in populations with heterogeneous immunocompromise. Our first application of expected utility theory answers: under which frequencies of environmental change is epigenetic regulation of gene expression likely to emerge? To answer this question, we propose that epigenetic modulation is itself a trait with benefits such as enabling rapid acclimation to environmental conditions and costs such as establishing and controlling epigenetic machinery. We then introduce the concept of epiallelic redundancy as a means by which to increase the likelihood that an organism will express the optimal trait for the current environmental conditions. Next, we develop a model characterizing the expected utility of epigenetic modulation of phenotype. Our results show that epigenetic control is only likely to evolve in situations where the cost of control is small and environmental fluctuations are frequent, causing variations in the fitness of a phenotype across a range of environmental conditions. Our next application of expected utility theory is to answer: can heterogeneity in signal detection criteria among individuals in a population facilitate or dissolve trade-offs in group decision making? To answer this question, we first invoke the framework of signal detection theory (SDT). In this framework, individuals vary in the criteria that they set to detect and respond to environmental signals. Individuals with high criterion values for signal detection have more hits (true positive detections) and fewer correct rejections (true negative detections), whereas individuals with lower criterion values will have fewer hits and more correct rejections. This difference in detection ability leads to trade-offs in signal detection. To answer our question at the population level, we expand the classical signal detection theory framework to include populations of individuals varying in their signal detection criterion choice. Further, we determine a net utility for each individual of participating in group consensus. We show that individual payoff for participating a population-level decision-making regarding the presence of a signal can affect the criterion choice utility each individual experiences. These results suggest that v coordination in decision-making regarding a signal can be advantageous in changing environmental conditions and can maintain individual variation in signal detection criteria in populations. Our final application of expected utility theory answers: how does heterogeneous immunocompromise within a population affect the economic and epidemiological utility of social isolation? To answer this question, we develop a multi-group Susceptible Exposed-Infected-Recovered (SEIR) model to compartmentalize individuals by immunocompetence. Next, we use expected utility theory to generate utilities of social isolation. These utilities of social isolation are then evaluated for various social isolation scenarios and provide a means by which to compare both macroeconomic and epidemiological effects of the social isolation scenarios. We show that in populations with high proportions of immunocompromised individuals, there is a higher expected utility of social isolation than in populations with smaller proportions of immunocompromised individuals. In populations with higher proportions of immunocompromised individuals, we find that the form of the expected utility curve is shifted such that more stringent social isolation is more favorable to economic and epidemiological outcomes. Taken together, the three chapters of this dissertation represent successful applications of expected utility theory to ecological systems. The conceptual and computational simplicity of expected utility theory lends itself readily to application in a wide variety of biological systems and provides a reliable proxy for quantifying the trade-offs inherent to those systems.
NotePh.D.
NoteIncludes bibliographical references
Genretheses, ETD doctoral
LanguageEnglish
CollectionSchool of Graduate Studies Electronic Theses and Dissertations
Organization NameRutgers, The State University of New Jersey
RightsThe author owns the copyright to this work.