An Emory College initiative working to cultivate conversations between experts in the life and physical sciences will hold a workshop Thursday, Jan. 21, to discuss how they can help get a handle on COVID-19.
Emory's interdisciplinary Theory and Modeling of Living Systems (TMLS) unites researchers in biology, computer science, physics, psychology and other fields, all of whom approach their work as a quantitative subject.
Workshops like the one this week help TMLS in its bid to become a national leader in cutting-edge research by talking in those disciplines’ common language: math.
Early COVID-19 modeling arrived at conflicting conclusions, in part because simple statistical approaches did not account for human behavior and biology, which affect transmission, and in part because more detailed mechanistic models meant to mimic transmission had too many moving parts to be adequately reflect the spread of the new virus the way they do with the less-contagious seasonal flu.
“We never had a virus spread like this, and it took a while to learn what to model and not model,” says Emory physics and biology professor Ilya Nemenman, who also serves as co-director of the TMLS.
“The main focus now is to take what we’ve learned, and to understand which features of human behavior and of how the virus interfaces with our immune system affect its spread the most, and hence must be accounted for in the models,” he adds.
“There will be a lot of such questions discussed at the workshop, aiming to understand how we can build models to guide specific quantitative decision about interventions,” Nemenman says.
Understanding biology with the precision previously reserved for physical sciences can be isolating in traditional academic environments. Emory, which has long encouraged interdisciplinary research from often-siloed researchers, established TMLS several years ago to formalize such efforts.
The initiative has strengthened Emory’s quantitative modeling, especially in behavioral biology, neurology, virology and public health. TMLS workshops create an opportunity for those scientists, both at Emory and globally, to interact and learn from one another.
They are technical affairs, historically drawing a few hundred scientists discussing how to build theoretical frameworks for understanding living systems.
When the pandemic moved the workshops online, attendance grew to more than 5,000 attendees participating in events, including previous workshops about neural motor control, material properties of biological tissues and animal behavior – and how to use machine learning to learn new laws of nature in these domains.
One workshop last hinted at the pandemic, asking how to make predictions in biological systems with the same accuracy required in physics.
This week’s symposium advances the discussion, asking what different fields’ see as necessary data and ideal models to predict COVID-19’s spread and design effective control strategies.
For Katia Koelle, the associate biology professor who co-directs TMLS, the workshop may help show the best quantitative ways to approach that data for real-world suggestions.
Koelle has long viewed biology as a quantitative science, applying statistical methods in her research to better understand patterns of evolution and adaption in seasonal influenza viruses and other human viruses. Those patterns in turn reveal the mechanisms the flu uses to escape immunity.
Combined with her specialty in the use of sequence data – sequencing the genetics of a viral sample – Koelle’s work may be applicable in understanding the rate of superspreading with the novel coronavirus and how many new viral introductions are occurring in a specific area.
Those conclusions influence public policy decisions on vaccination dosage and allocation.
“TMLS is a great hub to bring people together, with a nice synergy of expertise, for us to learn from one another,” Koelle says. “Understanding where we have gone wrong in categorizing patterns of spread – and where models have performed well – is critical for ongoing efforts aimed at controlling SARS-CoV-2.”