Though he may not have predicted he would be part of the modeling effort for a pandemic this year, Infectious Diseases Institute member Dr. Wasiur R. KhudaBukhsh‘s spot on the OSU/IDI COVID-19 modeling response team was all but inevitable.
Wasiur is a President’s Postdoctoral Scholar at the Mathematical Biosciences Institute. Wasiur has devoted much of his professional life to studying math and the ways it can be applied to real-world problems. As Wasiur’s disease modeling experiences have proven, the solutions math has the potential to provide are far from trivial.
Wasiur began his postdoctoral fellowship at OSU in October 2018, however, his journey at OSU started much earlier. While pursuing his PhD degree at the Technische Universität Darmstadt in Germany, Wasiur visited OSU’s Mathematical Biosciences Institute as part of a summer program. One of the people mentoring and showing Wasiur around the campus he would one day call home was current co-director of IDI’s Ecology, Epidemiology, and Population Health thematic program, Dr. Joe Tien.
Joe isn’t the only OSU modeler Wasiur interacted with before joining OSU himself—he worked with IDI member Dr. Greg Rempala quite substantially on various projects including one modeling the Ebola epidemic.
Wasiur, Greg, and Joe share a probabilistic network-based approach to mathematical modelling in epidemiology. Probabilistic models intentionally incorporate uncertainty in their predictions; they simulate a range of various scenarios using random variables and predict the likelihood of each scenario occurring. Probabilistic models are particularly useful for simulating future scenarios for which there is little or no historical data.
When COVID-19 emerged as a threat to Ohioans, the OSU/IDI COVID-19 modeling response team began forming with the goal of providing predictions to the Ohio Department of Health, Ohio Department of Medicaid, and Ohio Hospital Association—leading the team were Joe Tien and Greg Rempala. So, when a team of OSU experts was being assembled to model an unprecedented pandemic, it’s no surprise that Wasiur was tapped to join in the modeling effort.
Wasiur plays a critical role on the team. He is responsible for modeling the state-wide case occurrences and predicting potential hospital surges. Wasiur does this through use of the team’s dynamic network model, where the network is composed of interconnected Ohioans. The model uses the limited real data about confirmed cases in Ohio and abstracts the data to predict how the disease will spread across the network and, crucially to state decision-makers, how that spread will pressure Ohio’s healthcare system.
Real life is messy, though, and so are attempts to model it, as Wasiur can attest. Policy and behavioral changes in the real-life human network, such as lockdown measures and social distancing, have to be represented in the abstract one. Math, Wasiur says reverently, is abstraction to its core. A model, however, is more than a mathematical equation—it’s a representation of reality.
A good model, therefore, must find a balance between accounting for the complexity of reality without compromising the simplicity necessary for it to have wide practical application. Modeling, according to Wasiur, is all about maintaining this “fine balance” between fidelity to reality and prediction through abstraction.
Wasiur and the rest of the OSU/IDI COVID-19 modeling response team are continuing to fine tune and develop new ways to maintain this balance as the state begins to reopen. Wasiur says the team is continuing to grow as more experts are pulled into the effort to keep Ohio safe. Though his experience modeling COVID-19 is still far from over, Wasiur says his time on the team has already impacted his future research trajectory. It has emphasized the importance of epidemic modeling and raised new questions about public health problems. Fortunately for the public, Wasiur has the math and the modeling to solve them.
Written by Brooke Zentmeyer.
For more information about the dynamic network model used by the team, read their white paper, “Predicting COVID-19 Cases and Subsequent Hospital Burden in Ohio” and their blog post, “A note from the OSU/IDI COVID-19 Modeling Response Team.”
Dr. Wasiur R. KhudaBukhsh is a postdoctoral researcher in the Mathematical Biosciences Institute. Wasiur has a bachelor’s degree in statistics from the University of Calcutta, a master’s degree from the Indian Statistical Institute, and a PhD from the Department of Electrical Engineering and Information Technology at the Technische Universitat Darmstadt. His research interests include applications of probability theory and statistics to problems arising from epidemiology, biology, statistical physics, computer science, and engineering disciplines. When he’s not modeling novel viruses, Wasiur enjoys playing badminton, biking, and running on the Olentangy Bike Trail.