Amanda Perofsky is a Research Assistant Professor in the Department of Public Health and Health Sciences at Bouvé College, with joint appointments at the Network Science Institute, the Roux Institute, and the Institute for Experiential AI. Based at the Roux Institute in Portland, ME, her work focuses on understanding why respiratory viruses behave the way they do, and how we can get better at seeing outbreaks coming before they hit. She draws on a toolkit spanning statistics, mathematical modeling, and machine learning to study how viral evolution, population immunity, and human mobility shape the way diseases spread, and she puts that work into practice by contributing real-time forecasts to national efforts like CDC FluSight and the US Scenario Modeling Hub.
Amanda's path to respiratory viruses is anything but a straight line. She started out as a disease ecologist, earning her PhD from UT-Austin, where her research took her all the way to Madagascar to study how social networks shape the gut microbiomes of wild lemurs. She pivoted to respiratory epidemiology in 2018 during a postdoctoral fellowship at NIH's Fogarty International Center, later supporting pandemic response efforts — forecasting COVID-19 cases on U.S. military bases for the Department of Defense and partnering with South Africa's National Institute for Communicable Diseases. She then joined the Seattle Flu Study, where she explored how COVID-19 interventions and shifting mobility patterns disrupted the normal rhythms of viruses like influenza and RSV.
Outside the lab, Amanda spent over a decade as a college radio DJ at WUOG (UGA) and KVRX (UT-Austin), where she also co-hosted They Blinded Me with Science, a weekly science talk show. These days, her free time is filled with hiking, catching live music, and hanging out with her Australian cattle dog, Peewee. We're so glad to have her as part of the team!
Q1. Welcome to NetSI at the Roux Institute! What's one experiment or project you're excited to finally pursue now that you're here?
A. I'm excited to continue investigating how the COVID-19 pandemic disrupted the ecology of other respiratory viruses. In early 2020, stay-at-home orders, social distancing, and masking effectively eliminated the transmission of endemic viruses like influenza and RSV. As COVID-19 restrictions relaxed, their return to circulation was remarkably different depending on the type of virus: some viruses rebounded as soon as stay-at-home orders lifted while others took much longer to return to circulation. This provides a natural experiment to disentangle the relative roles of human behavior, population immunity, and environmental factors in shaping seasonal outbreaks. Maine is an exciting place to study these questions because detailed surveillance data exist through MaineHealth and Maine CDC, and it's a very different setting from the large urban metros that most studies focus on. Being part of NetSI with a community of experts in infectious disease modeling and human mobility is a great fit for this kind of work.
Q2. You've recently joined the founding task force for Northeastern's new Center for Public Health Technology. How does your research connect to that initiative?
A. The center is focused on bridging technological innovation and public health, which connects naturally to my work. I'm really interested in figuring out which data signals — whether from genomic surveillance, cellphone mobility data, or other sources — actually improve our ability to monitor and predict epidemics. Which datasets should we invest in collecting, which do we need in real time, and at what resolution? These are the kinds of questions the center is well positioned to address, and I'm looking forward to working with colleagues across disciplines to help shape that agenda.
Q3. What would success look like for your research program in 5 years?
A. In five years, I'd hope to have contributed to NetSI's growing efforts to bridge mechanistic understanding of disease dynamics with epidemic forecasting. I'd also like to return more to my earlier interest in seasonal influenza — there's still so much we don't understand about why flu seasons vary so dramatically from year to year. A lot of research on the various factors that shape flu epidemics is based on data collected before the 2009 H1N1 pandemic, and the evolutionary dynamics of H3N2 (the flu virus that evolves most rapidly and causes the most severe cases) have changed considerably since then. Genomic data in particular remain underutilized in forecasting flu epidemic dynamics, and I think there's a real opportunity there. I also want to make progress on understanding how children's contact and movement patterns shape respiratory virus spread — children bear the greatest burden of respiratory illness but are excluded from the large-scale behavioral datasets we typically use to study the spatial spread of infectious diseases. Studying respiratory virus dynamics in a predominantly rural state with an aging population and challenges around healthcare access could also provide insights that are more broadly relevant to underserved communities. Lastly, I'd love to be mentoring students and postdocs who are interested in the intersection of network science, computational modeling, and infectious disease epidemiology.
Q4. What do you do when you're stuck on a problem? Do you have a creative reset?
A. I go for a walk. In every city I've lived in, I've found a place I like to sit for a while. For example, in grad school it was the turtle pond on UT's campus, and in Seattle I lived at the top of Queen Anne Hill close to Marshall Park, which has stunning views of the Olympic Mountains. I'm still finding my spot in Portland, and it’s very cold right now, but the Roux Institute is close to the Eastern Promenade so that's a strong contender.
