Mathematical models have been used throughout the COVID-19 pandemic to better understand and predict patterns of viral spread both at the human population level and at the within-host level. Here, I will first talk about recent work from my group that uses branching processes and within-host ODE modeling to ask whether non-pharmaceutical interventions such as masking have the potential to impact the risk of developing severe disease. I will follow up on this work to show how branching process models can be interfaced with viral sequence data to statistically estimate transmission bottleneck sizes, defined as the number of viral particles that establish infection. Together, our results indicate that very few (<5 viral particles) generally are responsible for establishing infection and point towards masks not being useful for reducing the risk of developing severe disease, conditional on infection (although clearly useful in reducing the risk of infection).
Department of Mathematics