Computer Science ETDs

Publication Date



How different is the immune system in a human from that of a mouse? Do pathogens replicate at the same rate in different species? Answers to these questions have impact on human health since multi-host pathogens that jump from animals to humans affect millions worldwide. It is not known how rates of immune response and viral dynamics vary from species to species and how they depend on species body size. Metabolic scaling theory predicts that intracellular processes are slower in larger animals since cellular metabolic rates are slower. We test how rates of pathogenesis and immune system response rates depend on species body size. We hypothesize that immune response rates are invariant with body size. Our work suggests how the physical architecture of the immune system and chemical signals within it may lead to nearly scale-invariant immune search and response. We fit mathematical models to experimental West Nile Virus (WNV, a multi-host pathogen) infection data and investigate how model parameters characterizing the pathogen and the immune response change with respect to animal mass. Phylogeny also affects pathogenesis and immune response. We use a hierarchical Bayesian model, that incorporates phylogeny, to test hypotheses about the role of mass and phylogeny on pathogen replication and immune response. We observe that: 1) Hierarchical models (informed by phylogeny) make more accurate predictions of experimental data and more realistic estimates of biologically relevant parameters characterizing WNV infection. 2) Rates of WNV production decline with species body mass, modified by a phylogenetic influence. Our work is the first to systematically explore the role of host body mass in pathogenesis using mathematical models and empirical data. We investigate the complex interplay between the physical structure of the immune system and host body mass in determining immune response. The modeling strategies and tools outlined here are likely to be applicable to modeling of other multi-host pathogens. This work could also be extended to understand how drug and vaccine efficacy differ in humans from model organisms like mice where most immunological experiments are conducted.




scaling; immune system; mathematical modeling; viral dynamics; lymph node scaling

Document Type


Degree Name

Computer Science

Level of Degree


Department Name

Department of Computer Science

First Advisor

Moses, Melanie

First Committee Member (Chair)

Forrest, Stephanie

Second Committee Member

Perelson, Alan

Third Committee Member

Koster, Frederick

Fourth Committee Member

Lane, Terran