Jonathan Gilligan helps fight COVID-19 in Latin America
Vanderbilt’s College of Arts and Science has long had connections to Latin America. Through the Center for Latin American Studies, the work of Latin American specialists in departments such as history and anthropology, and numerous student and faculty research trips, the college has forged strong ties to the region, its universities, and its people. So as the novel coronavirus began to spread around the world, some Arts and Science faculty turned their attention south, to see how they could help communities and colleagues in Latin America.
Associate Professor of Earth and Environmental Sciences Jonathan Gilligan pivoted from his usual activities to work with Latin American epidemiologists on predicting and preventing the spread of COVID-19. A specialist in interactions between human society and the environment, Gilligan often collaborates with sociologists and psychologists to study why people make certain choices related to climate change, as well as how they react and adapt to environmental stress. In some of his research, he uses computational modeling to calculate how people’s individual choices add up to affect society as a whole.
One of Gilligan’s graduate students, the son of a Colombian epidemiologist, recognized agent-based modeling’s potential to fight the spread of COVID-19. He shared Gilligan’s work with his father, who in turn told his colleagues about it and contacted Gilligan for help.“It’s called agent-based modeling, and you can also apply it to animals: how birds form flocks or fish form schools. It’s very applicable to the spread of epidemics because it can model how people spread a disease by interacting with others in their social networks,” Gilligan said.
Gilligan and his lab started by customizing their agent-based model for the Colombian situation. A first version made accurate calculations but was extremely slow. Gilligan was able to revise the model to make calculations for millions of people, and he and the Colombian epidemiologists are now calibrating it using real-life contact data.
The group is using the model to investigate some of the Colombian epidemiologists’ most interesting findings. Most Colombian infections appear to have started with people who were working abroad and brought the disease home with them. The disease then spread along their social networks—but not necessarily in the way Gilligan and his collaborators expected. In particular, they have found that occupation may be more important than number of contacts, since people who intersect with the health care system appear to spread COVID at the highest rate.
Ultimately, Gilligan believes that his model will prove useful on two fronts. First, it can help Colombian epidemiologists predict the locations of COVID outbreaks, so they can shore up resources in advance and prevent local health care systems from becoming overwhelmed. Second, they can use the model to test possible interventions, such as varying degrees of social isolation, and determine which are mostly likely to slow the spread of COVID
Two weeks after Gilligan began working with the Colombian epidemiologists, Vanderbilt University Medical Center contacted him about adapting his model for Tennessee. He also started discussing possible projects with colleagues in the university’s Data Science Institute and the Department of Psychology. The research could explore synergies between Gilligan’s models, artificial intelligence, and neuroscience modeling, with results used to develop solutions for fighting future pandemics. All these projects, he says, are excellent examples of how his position in the College of Arts and Science enables world-changing interdisciplinary work.
“I don’t have a background in epidemiology at all. I couldn’t contribute to the pandemic fight if I didn’t have the opportunity to work with experts outside my field,” Gilligan said. “In turn, I can share what I know about modeling—each person can share information and say, ‘I don’t know about this. Can you explain it to me?’ This is where Vanderbilt’s support for interdisciplinary work is really helpful. It’s critical.”