Predicting Zoonotic Pandemics

Colorized scanning electron micrograph of a cultured animal cell (blue) heavily infected with SARS-COV-2 virus particles (orange), isolated from a patient sample.
Colorized scanning electron micrograph of a cultured animal cell (blue) heavily infected with SARS-COV-2 virus particles (orange), isolated from a patient sample. Image captured and color-enhanced at the NIAID Integrated Research Facility (IRF) in Fort Detrick, Maryland. Credit: NIAID

One legacy of COVID-19 may be an increasing number of future investigations into wildlife virology and other pathogens with the goal of predicting the seeds of the next zoonotic-based pandemic. University of Florida associate professor of medical geography Sadie Ryan recently collaborated on two studies focused on tools and models that can be employed to better predict and prevent zoonotic epidemics. Ryan is a medical geographer in the UF College of Liberal Arts and Sciences and a faculty member of the UF Emerging Pathogens Institute.

Headshot of Dr. Sadie Ryan.
Headshot of Dr. Sadie Ryan.

The first paper, in Nature Microbiology, proposes a network science framework to understand and predict the susceptibility of people and animals to viral infections. The second paper, in Philosophical Transactions of the Royal Society B, explores ethical and practical questions about how future technologies may be used in predicting zoonotic risks.

The first paper reviews models used to predict zoonotic epidemics, but it notes that “spillover risk will never be reduced to zero.” And, after a new virus gets into a human, the tools for preventing a pandemic rest with diagnostic and surveillance efforts, healthcare access and social safety nets. Therefore, predictive modeling tools need to focus upstream of spillovers.

The team created a host-virus model database that organizes the past decade of predictive zoonotic risk modelling studies. But the biggest limitation of using models to predict zoonotic epidemic risk is a lack of data stemming from viral discovery being in its infancy. Only about 1% of mammal viruses have been discovered, they note; a number that further decreases for all other vertebrates.

In the second paper, which was written by largely the same group of collaborators, the authors turn their attention to future technologies that may be developed to predict zoonotic risks. They define the term zoonotic risk technologies as a family of approaches to identify viruses that could potentially become zoonoses. According to the study, researchers will likely use machine learning and other data-driven techniques to sort through massive amounts of zoonotic data in search of leads as to which wildlife pathogens may pose future a threat to people.

These technologies tend to assume that future or undetected zoonoses are more similar to those that have spilled over before them than those viruses that have not. However, the authors predict that in time, genomic approaches will likely move beyond identifying genetically similar wildlife pathogens with pandemic potential and begin to predict which viruses are compatible with human cells.  But even this approach may be limited by a lack of data: there are only a hundred or so known zoonoses which could prove limiting.

Chart showcasing host-virus associations.
Established host-virus associations can be used to predict transmission that crosses the species barrier.

The authors explore how the rise of zoonotic risk technologies demands a close examination of questions regarding how these techniques should be developed and applied. Other issues to consider are the effect of these technologies on global health, who controls and accesses them, who benefits from them, whether they can improve pandemic prevention, and whether they may create new or unforeseen challenges.

Predictions made by computational zoonotic risk technologies would also need to be tested in a real world laboratory. Experimental data could be used to probe molecular-level barriers to zoonotic emergence. But the authors acknowledge that simply knowing about the potential of a zoonose is not always enough to influence helpful public health outcomes due to barriers in the pipeline that translate theory or academic findings into applied actions.

The authors also ponder the downside of zoonotic risk technologies:

  • Unless the benefits of new health or zoonotic risk technologies are equally shared, a global injustice unfolds
  • They could be used to create new viral sequences or even synthetic viruses with a high potential for seeding epidemics or pandemics
  • The majority of viruses with zoonotic potential reside in tropical countries, however researchers in North American and Europe are most likely to benefit from zoonotic risk technologies via career advancement
  • Data scarcity is a large limiting factor

The authors acknoeldge that ever greater prediction abilities do not equate to increased prevention. Models will never perfectly represent or predict reality, and as this field grows the limitations and risks of its findings need to be carefully communicated.

As climate change continues, researchers predict there will be more occurrences of zoonotic spillovers. And as researchers better understand the processes and dynamics that lead to spillovers, human-to-human transmission, epidemics, and pandemics, the field of zoonotic risk technology is sure to grow.

Acknowledgments: The work communicated here is part of an ongoing collaboration called Verena, visit to learn more. Verena collaborators are dedicated to open science and making available all the projects, codes, and databases that they work on. “When we publish, our end goals are not really the papers,” Ryan said. “They’re the means to communicate our progress and what we’re about, but it’s more of a holistic collaboration, with massive effort on architecting data and databases, and scoping how it all interfaces with existing frameworks, or forces us to create new ones.”

Authors of the Nature Microbiology paper: Gregory F. Albery, Daniel J. Becker, Liam Brierley, Cara E. Brook, Rebecca C. Christofferson, Lily E. Cohen, Tad A. Dallas, Evan A. Eskew, Anna Fagre, Maxwell J. Farrell, Emma Glennon, Sarah Guth, Maxwell B. Joseph, Nardus Mollentze, Benjamin A. Neely, Timothée Poisot, Angela L. Rasmussen, Sadie J. Ryan, Stephanie Seifert, Anna R. Sjodin, Erin M. Sorrell, Colin J. Carlson 

Authors of the Philosophical Transactions of the Royal Society B paper: Colin J. Carlson, Maxwell J. Farrell, Zoe Grange, Barbara A. Han, Nardus Mollentze, Alexandra L. Phelan, Angela L. Rasmussen, Gregory F. Albery, Bernard Bett, David M. Brett-Major, Lily E. Cohen, Tad Dallas, Evan A. Eskew, Anna C. Fagre, Kristian M. Forbes, Rory Gibb, Sam Halabi, Charlotte C. Hammer, Rebecca Katz, Jason Kindrachuk, Renata L. Muylaert, Felicia B. Nutter, Joseph Ogola, Kevin J. Olival, Michelle Rourke, Sadie J. Ryan, Noam Ross, Stephanie N. Seifert , Tarja Sironen, Claire J. Standley, Kishana Taylor, Marietjie Venter, and Paul W. Webala

Written by: DeLene Beeland