Home / News / Predicting future pandemic patterns

Predicting future pandemic patterns

Since 2020, the coronavirus has infected more than 627 million people and caused more than 6.5 million to lose their lives, according to the World Health Organization. COVID-19 has put a strain on almost every country in the world, fundamentally altering their economies and health care systems worldwide.

Professor Pavan Turaga, an expert in machine learning, is leading an interdisciplinary research team investigating the use of machine learning to reduce the societal impacts of future pandemics. Photo courtesy Unsplash

A study published in the Proceedings of the National Academy of Sciences that analyzed data from pandemics dating back to 1600 reports that future pandemics caused by illnesses like the coronavirus are expected to arise more frequently. The study estimates the yearly likelihood of severe disease outbreaks could increase by 300% in the decades to come.

With the growing threat looming, a team of 17 researchers from Arizona State University is looking to minimize the impact of these future pandemics through the power of machine learning.

Led by machine learning expert Pavan Turaga, director of the School of Arts, Media and Engineering in the Herberger Institute for Design and the Arts and a professor in the School of Electrical, Computer and Energy Engineering, part of the Ira A. Fulton Schools of Engineering at ASU, the team brings together computational, biological and social science experts to develop modeling tools that can adapt to predict the spread of both new and existing pathogens to inform legislative and health care responses to pandemics.

The team has earned one of only 26 grants from the National Science Foundation’s (NSF's) Predictive Intelligence for Pandemic Prevention program to move their project forward.

Turaga says current mathematical models used to predict the spread of pathogens are based on designs from decades ago that don’t account for nuances such as socioeconomic conditions and climatic factors like extreme heat. These factors are also known as compound effects.

“We still do not have a trustworthy model that can predict any pandemic spread reliably,” Turaga says. “The grant is not to develop another narrow model, but instead to do a ‘scan’ of what factors are not being currently considered, and what new techniques can be leveraged over what has been used earlier.”

Collaborating to conquer pandemics through research

Initially, the researchers comprised teams working on projects funded by three different grants to study and model the spread of the coronavirus under the NSF’s Grants for Rapid Response Research program in 2020. The research teams kept in contact with each other, found out about the Predictive Intelligence Pandemic Prevention program and came together under the new 18-month research grant as one large, interdisciplinary team, led by Turaga and four co-principal investigators.

Gautam Dasarathy, an assistant professor of electrical engineering, is a co-principal investigator leading efforts in machine learning for pandemic spread modeling.

“I am extremely honored and thrilled to be part of this research,” Dasarathy says. “It brings together a truly diverse team of researchers with expertise ranging from virology, sequencing and wastewater analysis to data science, controls and cryptography, to name a few.”

Another co-principal investigator, Patricia Solis, executive director of Knowledge Exchange for Resilience at ASU’s School of Geographical Sciences and Urban Planning, leads a team looking at how human behavioral and socioeconomic factors affect pandemic spread models. Solis’ goal is to connect with community leaders making pandemic response decisions and to determine what criteria have been taken into account for both the pandemic phase of COVID-19 and the phase in which the world learns to live with the disease.

“What kinds of data, models, information and insights are being used, and what else is needed for better decision-making?” Solis asks. “We aim to bring this back to the conversations with our ASU scientists about how to design better models that are not only robust, but also truly serve research of public value, as per our ASU Charter.”

Giulia Pedrielli, another co-principal investigator and an associate professor of computer science and industrial engineering in the School of Computing and Augmented Intelligence, a part of the Fulton Schools, is lending her expertise in sensing mobile activity while respecting users’ privacy to determine how best to allocate pandemic response resources.

Pedrielli believes researchers’ consideration of additional factors to inform pandemic response decisions will lead to better outcomes for society.

“We hope including behavior in testing and resource allocation decisions will make distribution of resources more equitable, considering multiple and diverse aspects of the effect of policies not only on mechanics of pandemic spread, but also on behavior,” she says.

Visar Berisha, a Fulton Schools associate professor of electrical engineering with a dual appointment in the College of Health Solutions, is a co-principal investigator leading a team that uses machine learning to analyze patterns in numerical parameters using biological samples taken from blood, saliva and wastewater.

Berisha hopes the biological sample research can be used for an easily accessible dashboard that will constantly look out for new pathogens, such as new variants of the coronavirus or entirely new diseases.

“Based on different regions in the U.S. with different socioeconomic cohorts, the idea is to develop on-the-fly precision interventions to inform people of the risks of these pathogens,” he says. “By providing actionable information about what they can do to protect themselves and those around them, we can reduce spread and improve outcomes.”

This actionable information could include public health advice such as encouraging people with compromised immune systems or who work in high-risk settings to wear masks, or advising those most vulnerable to COVID-19’s worst effects to get additional booster shots.

With a little help from research friends

While only ASU researchers will be initially involved in the project, the team aims to forge partnerships with outside organizations to assist its efforts. Those include Creative Testing Solutions, which analyzes blood samples collected from the ASU community by Berisha’s research group, and Cowper Sciences Inc., which analyzes biological samples to determine how the immune system responds to pathogens.

In addition to the team of ASU experts, students will have opportunities to gain hands-on research experience by assisting with the project.

Berisha says the project demonstrates the benefit of basing the work at ASU because of the variety of expertise and the range and depth of inquiry its researchers can contribute to the endeavor.

“You have people wearing very different hats on an everyday basis that are now coming together under the same roof to work on a very important problem,” he says.

In addition to Pavan Turaga, Visar Berisha, Giulia Pedrielli, Gautam Dasarathy and Patricia Solis, the research group includes other Fulton Schools researchers: Daniel Rivera, a professor of chemical engineering in the School for Engineering of Matter, Transport and EnergyRolf Halden, director of the Biodesign Center for Environmental Health Engineering and a professor of civil and environmental engineering in the School of Sustainable Engineering and the Built EnvironmentLalitha Sankar, an associate professor of electrical engineering in the School of Electrical, Computer and Energy Engineering; and Ni Trieu, an assistant professor of computer science in the School of Computing and Augmented Intelligence.

The group also includes Erik Johnston, the interim deputy director of and a professor in ASU’s School for the Future of Innovation in SocietyMichael Simeone, an associate research professor in ASU’s School of Sustainability and the Global Biosocial Complexity InitiativeTimothy Lant, the director of program development for ASU’s Knowledge EnterpriseWenwen Li, a professor of geographic information science in ASU’s School of Geographical Sciences and Urban Planning; Laimonas Kelbauskas, a research scientist in ASU’s Knowledge Enterprise; Efrem Lim, an associate professor of microbiology in ASU’s School of Life SciencesMatthew Buman, professor and director of ASU’s College of Health Solutions; and Neal Woodbury, vice president and chief science and technology officer for ASU’s Knowledge Enterprise.

tjtriolo@asu.edu