Skip to content

Back to Departmental Colloquium: Fall 2025

Departmental Colloquium


Date: Thursday, Nov 20, 2025

Time: 4:00PM - 5:00PM

Location: JWB 335


Amber Smith

University of Tennessee Health Science Center

Title

Uncovering Infection Dynamics through Parameter Estimation and Virtual Patients

Abstract

Understanding infection dynamics requires mathematical tools that can translate varying degrees of experimental and clinical data into meaningful biological insight. I will discuss approaches for estimating key parameters in within-host models of respiratory infections using simulated annealing and bootstrap techniques to fit infection data, characterize parameter behavior and uncertainty, and extract clinically relevant information. When empirical data are limited, virtual patient cohorts offer a way to explore clinically relevant variability and identify drivers of divergent infection outcomes. These analyses show that distinct parameter sets and trajectories do not necessarily produce individualized outcomes. Because existing strategies for generating virtual patients can be computationally inefficient, I will outline current approaches and highlight the potential of trajectory-matching (TM)-ABC-MCMC to improve efficiency while preserving biological realism. Together, these methods show how parameter-informed modeling can deepen our understanding of respiratory infection dynamics and support the development of predictive tools.

Applied Mathematics Mathematical Biology

Calendar file