Computational Epidemiology of Large Multi-Dimensional Data Streams in Malaria and COVID-19
Open Access
- Author:
- Tran, Thu
- Graduate Program:
- Biology
- Degree:
- Doctor of Philosophy
- Document Type:
- Dissertation
- Date of Defense:
- May 10, 2023
- Committee Members:
- Stephen Schaeffer, Co-Chair of Committee
Justin Pritchard, Co-Chair of Committee
Manuel Llinas, Outside Unit & Field Member
Elizabeth Mcgraw, Program Head/Chair
Maciej Boni, Major Field Member & Dissertation Advisor - Keywords:
- malaria
plasmodium
falciparum
mathematical model
public health
epidemiology
fitness
evolution
infectious disease
antimalarial
kelch13
sub sahara africa
bayesian
rhode island
massachusetts
connecticut
drug resistance
vaccine
vaccination strategy
optimization
population immunity
attack rate
pandemic
USA
doubling time
covid 19
sars cov 2
differential equation
artemisinin - Abstract:
- Mathematical modeling plays a key role in helping us understand disease dynamics, answer counterfactual questions, and evaluating intervention strategies. This dissertation focuses on the utilization of computational tools to tackle challenges in public health, specifically in controlling and managing two of the high-mortality infectious diseases, Plasmodium falciparum malaria and COVID-19. I present an approach to estimate antimalarial drug efficacies for unobserved Plasmodium falciparum genotypes. This drug-by-genotype table allows mathematical models of drug resistance evolution to simulate the competition among any P.falciparum genotypes in any possible combination of available antimalarial drugs. I then use these estimates along with survey data on treatment-seeking behavior and drug choice to investigate the doubling time of kelch13 resistant genotypes in Sub-Saharan Africa. Kelch13 mutations which confer resistance to artemisinin have emerged as a major concern, threatening the efficacy of artemisinin-based combination therapies, especially in high malaria burden Sub-Sahara African countries. By assessing the risk of kelch13 variants being imported and selected for in this region, we can then explore drug resistance management strategies to slow their spread while increasing treatment access. The second half of this dissertation focuses on COVID-19. I evaluate various SARS-CoV-2 vaccine allocations to determine the optimal strategy to prevent the most number of COVID-19 deaths and hospitalizations for Rhode Island and Massachusetts under low, medium, and high transmission scenarios. In the last research chapter, I show how we can harness rich multi-dimensional data streams using mathematical model and statistical inference to estimate SARS-CoV-2 population immunity and attack rates in three southern New England states of the US. Being able to monitor these key epidemiological quantities in real-time will provide us with valuable insights into the population's susceptibility and help evaluate the effectiveness of various public health interventions.