Southern bentwing bat (Steve Bourne, Wikimedia Commons)
10 November 2017. A team of ecologists and mathematicians is applying machine learning and statistical models to forecast outbreaks of zoonotic diseases, those spread to humans from animals. The five-year research project is funded by a $2 million grant from National Science Foundation.
Researchers from Cary Institute of of Ecosystem Studies in Millbrook, New York, University of Georgia in Athens, and North Carolina A&T State University in Greensboro aim to develop better tools for predicting outbreaks of zoonotic diseases, which resulted in several recent epidemics among human populations, such as Ebola, Zika, and severe acute respiratory syndrome, or SARS. In many cases, humans acquire the viruses or bacteria from animals, such as bats or rodents, but also livestock.
Barbara Han, a disease ecologist with the Cary Institute and project leader, says the team’s research expects to give health authorities better methods to reveal and prevent upcoming threats from zoonotic disease. “We want to help shift society from a reactive to a proactive approach to managing zoonotic disease,” says Han in a Cary Institute statement. “Instead of responding to outbreaks, let’s try to stop them from happening in the first place.”
Han and colleagues are using machine learning to analyze large databases of mammals, pathogens associated with those animals, and environmental conditions to distill the properties and characteristics associated with the occurrence and spread of zoonotic diseases. These analyses will then be the basis for statistical models to map locations and conditions for future zoonotic disease outbreaks.
From this work, the researchers expect to identify traits of animal species and pathogens, as well as environmental conditions, that are the best predictors of zoonotic disease patterns. The team also plans to uncover factors in the environments of animals and pathogens associated with human infections from zoonotic diseases that can help predict the extent of outbreaks in human populations. The researchers anticipate applying statistical models to reflect these patterns of animal and human infections, and provide evidence-backed tools for predicting zoonotic disease outbreaks based on properties of animal hosts, pathogens, and environmental conditions.
Suzanne O’Regan, a mathematics professor at North Carolina A&T and a researcher on the project, notes that “Over 50 life history features are being incorporated into models for most mammal groups.” Data being collected about mammals include physical, metabolic, and reproductive characteristics, but also timing of the animals’ activities: day, night, dawn, or dusk. Pathogens are being classified on their transmission mechanisms, survivability in hosts or environments, and if they can sustain transmission of disease to many people, or only one person.
Han notes that their project brings advanced data analytics together with classic biological investigations. “These tools merge data mining and machine learning with established methods of studying disease dynamics,” says Han, “to help us think carefully about what’s distinguishing animal groups from each other in terms of zoonotic disease, and eventually, for risk of human spillover and epidemics.”
More from Science & Enterprise:
- Big Data Analysis Platform for Brainwaves in Development
- IBM, UC San Diego Partner on Healthy Aging
- Algorithms ID Factors in Parkinson’s Progression
- Geisinger, Pharmas Partner on Diabetes Model
- Algorithms, Imaging Help Predict Alzheimer’s Onset
* * *
RSS - Posts
You must be logged in to post a comment.