Research Interests
The role of human mobility in respiratory virus epidemics
During the COVID-19 pandemic, mobile phone location data became crucial for tracking population movements and evaluating control strategies against SARS-CoV-2 spread. I began working with cellphone mobility data in 2020, first to forecast COVID-19 outbreaks on military bases for US DoD, and later to analyze relationships between population mobility and healthcare-seeking behavior in South Africa and respiratory virus transmission in Seattle.
Mobility behavior and respiratory virus transmission dynamics in the Seattle metropolitan area
Mobility data have been widely used to model SARS-CoV-2 dynamics, but relationships between mobility behavior and the transmission of endemic respiratory pathogens are less understood. As a research scientist for the Seattle Flu Alliance, I investigated how mobility behavior, masking, and COVID-19 non-pharmaceutical interventions (NPIs) influenced the transmission dynamics of SARS-CoV-2 and 17 endemic respiratory viruses in the Seattle metropolitan area over a four-year period, 2018 - 2022 (Perofsky et al. 2024 Nat. Commun.). Building on this work, we are using mechanistic models to study the effects of waning immunity and decreased social distancing on the post-pandemic rebound of RSV in Seattle (medRxiv preprint). Additionally, I have collaborated with Trevor Bedford’s group (Fred Hutch) to link cellphone mobility patterns to the genomic epidemiology of SARS-CoV-2 in Seattle-King County and Washington state.
Ecological and evolutionary drivers of influenza outbreaks in human and animal populations
The effects of antigenic drift and subtype interference on annual influenza epidemics in the United States
Collaborators: Cécile Viboud (NIH), John Huddleston (Fred Hutch), Trevor Bedford (Fred Hutch), Florian Krammer (Mt. Sinai), WHO Global Influenza Surveillance and Response System (GISRS) Collaborating Centers in Atlanta, London, Melbourne, and Tokyo
Influenza viruses continually accumulate genetic changes in epitopes of two major surface proteins, hemagglutinin (HA) and neuraminidase (NA), in a process called “antigenic drift.” Antigenic drift enables the virus to evade immune recognition, leaving previously exposed individuals susceptible to reinfection and requiring regular updates to the flu vaccine. Antigenic drift is expected to increase susceptibility, leading to more flu cases and in turn earlier or more severe epidemics. However, epidemiological evidence for the impact of antigenic drift on seasonal outbreaks is conflicting.
In a recent eLife article (Perofsky et al. 2024), we compared experimental and sequence-based measures of A(H3N2) virus evolution in predicting regional epidemic dynamics in the US across 22 seasons (1997-2019), while also accounting for co-circulating flu viruses, prior population immunity, and vaccine-related factors. We found that evolution in both major surface proteins contributes to variability in epidemic magnitude across seasons, although viral evolution appears to be secondary to subtype interference in shaping annual outbreaks. When comparing the predictive performance of different evolutionary indicators, genetic changes in broad sets of epitope sites had stronger, more consistent relationships with A(H3N2) epidemic dynamics than the serological assays traditionally used to measure antigenic changes. We are now studying the effects of antigenic drift and prior immunity on A(H3N2) vaccine effectiveness in the US.
Influenza transmission dynamics among exhibition swine in the United States Midwest
Collaborators: Martha Nelson (NIH) and Andrew Bowman’s group (Ohio State)
Each summer in the US, youths attending agricultural fairs are exposed to genetically diverse influenza A viruses (IAVs) circulating in exhibition swine, leading to hundreds of zoonotic infections since 2010. Exhibition swine, comprising 1.5% of the U.S. herd, present a targeted opportunity to mitigate a pandemic threat. We analyzed virologic surveillance data from thousands of pigs at 350 national, state, and local swine exhibitions across several states during 2016–2018.
Key findings:
- An early-season national show played a key role in the propagation and spread of a specific virus (H1 δ-2) among exhibition swine, leading to the majority of zoonotic infections in the US during 2018 (Nelson, Perofsky, et al. 2020 J. Virol.).
- The earlier timing of jackpot shows and long-distance travel of pigs for repeated showing provide a pathway for the introduction of influenza into county fairs (McBride† and Perofsky† et al. 2021). Unlike state and county fairs, jackpot shows focus solely on swine and occur at the start of the show season.
Operational involvement in the COVID-19 pandemic response and disease predictions
- 2023 – , Dashboard of SARS-CoV-2 variant forecasts for Washington and other US states, hosted by the Seattle Flu Alliance (In collaboration with Nextstrain)
- 2022 – , Contributor to the US CDC FluSight Forecasting Collaboration. Submitted weekly short-term forecasts of influenza hospitalizations during the 2022-23 and 2023-24 seasons.
- 2022 – , Contributor to the US Influenza Scenario Modeling Hub. Submitted long-term scenario projections of influenza hospitalizations during the 2022-23, 2023-24, and 2024-25 seasons.
- 2020 – 2022, Developed the analysis, wrote the first report, and provided technical support to South Africa’s National Institute for Communicable Diseases for their COVID-19 Private Consultations Excess Respiratory Encounters Reports. Reports tracked excess respiratory encounters at hospitals, emergency departments, and primary care providers across different age groups and provinces. Bi-weekly and monthly reports were updated from September 2020 to April 2022.
- 2019 – 2022, Contributor to the US Department of Defense COVID + ILI Forecasting Collaboration. Submitted weekly short-term forecasts of influenza-like illness and COVID-like illness cases on US military bases during the 2019-20, 2020-21, and 2021-22 seasons.
PhD research: Ecological, evolutionary, and behavioral determinants of gut microbiomes in Malagasy mammals.
Advisor: Lauren Ancel Meyers (UT-Austin); Collaborators: Rebecca Lewis (UT-Austin), Anthony Di Fiore (UT-Austin)
Mammalian gut microbial communities govern host development, metabolism, immune function, and physiology, through interactions that range from commensal and mutualistic to pathogenic. However, the relative contributions of vertical, horizontal, and environmental transmission to gut microbiome composition are not well understood. For my dissertation research, I integrated field-collected data, molecular analyses, and computational and statistical approaches to study how social networks and proximity to other host species influence bacteria transmission and gut microbiome composition among the wild mammals inhabiting Kirindy Mitea National Park in western Madagascar. This research was conducted as part of The Sifaka Research Project, a long-term research study led by Dr. Rebecca Lewis (UT-Austin).
Key findings:
- In a single season, the gut microbiomes of wild Verreaux’s sifaka (Propithecus verreauxi) clearly reflect their social group membership, and both grooming and scent-marking behaviors promote microbial exchange and within-host diversity (Perofsky et al. 2017 Proc. R. Soc. B.).
- Sifaka social groups harbor distinct gut microbial communities over the course of several years, and compositional changes in the microbiota of individual animals are influenced by the gain or loss of unique social partners and male dispersal between groups (Perofsky et al. 2021 Mol. Ecol.).
- The ecological relationships among mammalian gut microbiomes mirror their hosts’ phylogeny. However, the predicted functionality of lemur microbiomes differentiate according to diet, and distantly-related terrestrial mammals have overlapping microbial communities, suggesting that ground dwelling facilitates the indirect horizontal transmission of gut bacteria among sympatric wild mammals (Perofsky et al. 2018 ISMEJ).
Together, these findings demonstrate that patterns of gut microbiome composition in wild mammals are scale-dependent: host phylogeny, diet, and substrate use shape microbial variation among sympatric mammal taxa, while social groupings and social contacts constrain the horizontal transmission of gut bacteria within a single host population.
Predoctoral Research
I graduated from the University of Georgia in 2009 with degrees in Biology and Ecology. As a UGA CURO Scholar and NSF REU awardee, my undergraduate research focused on salamander ecology and population modeling (Maerz Lab, UGA). In Fall 2009, I worked as a research assistant in Andrew Park’s Lab (UGA), studying environmental drivers of hemorrhagic disease virus outbreaks in white-tailed deer. During 2010-2011, I completed an NIH postbac fellowship, where I characterized viral infection profiles in Sjögren’s Syndrome patients (Adeno-Associated Virus Biology Section, NIDCR).