Apply to KAUST
Moving to KAUST
Search for Jobs
As countries around the world work to "flatten the curve," a reference to slowing the progression of COVID-19 cases, David Ketcheson is putting mathematical modeling to work to understand how transmission is playing out.
"I study things like waves," Ketcheson said. "Of course waves are made up of many tiny water molecules, and we cannot predict the movement of each of those molecules. But we can very accurately predict the motion of the whole wave."
Ketcheson describes the evolution of highly complex systems through differential equations, something he brought to bear to help explain the progression of SARS-CoV-2, the strain of coronavirus currently infecting hundreds of thousands worldwide, resulting in the illness called COVID-19.
We spoke with David Ketcheson for episode 10 of Sciencetown - Understanding the pandemic.
Ketcheson is an associate professor of applied mathematics and computational science at KAUST. He is quick to explain that he is not an epidemiologist. But, with historical data and mathematical heft, he does have some powerful points to make about the current COVID-19 situation.
"So, while we can't predict the actions of individual people, in a large population, we can pretty accurately predict the overall spread," Ketcheson said.
In late February, he found himself explaining to relatives and colleagues how what is now a global pandemic might progress. By mid-March, he had written a series of articles explaining, in layman's terms, the SIR model—a classical differential equation that breaks people into three groups, Susceptible, Infected and Recovering, for the purposes of modeling the spread of infectious disease.
What Ketcheson explained in those articles is rapidly playing out—the disease is spreading at an exponential rate. Perhaps more alarmingly, there might be as many as between 5 and 10 times more cases than are currently being reported. At a time when Johns Hopkins is reporting that over 700,000 cases have been confirmed worldwide, Ketcheson sees a world where between 3 and 6 million are most likely already infected, even if many of those individuals are not showing symptoms.
"The model has been powerfully predictive. In the absence of interventions, the doubling time of the disease would be between three and four days," Ketcheson said.
Early published estimates of the basic reproduction number for COVID-19 were around 2 to 2.5, but Ketcheson's analysis found the data to be consistent with a reproduction number as high as 5. More recent analyses by other experts also have consistently suggested that the number is significantly higher than originally thought.
The reproduction rate is the number of people, on average, that an infected person will pass on the virus to—further demonstrating the need for social distancing, and in the case of high-risk individuals, isolation.
"Flattening the curve is all about reducing that reproduction number, reducing the average number of individuals that people come in contact with when they are sick," Ketcheson said. "And, by reducing that, we slow down the exponential growth, so it takes longer to reach the peak, but we also reduce the height of that peak."The individual molecules in Ketcheson's wave are vitally important—they are friends, colleagues and loved ones. But the individual experiences of people become less helpful as we try to gain clarity about how big a wave COVID-19 is likely to be and how long it might take before we start to come down from the peak.
Ketcheson is part of a global academic collaboration that seeks to make powerful data accessible while encouraging happy accidents—the kinds of discoveries that come from highly multidisciplinary environments where users turn disparate kinds of data into powerful insights. The team, made up of KAUST faculty from marine science, computer science, mathematics, genomics and more, looks to publish a dashboard that does both.
One example might be the combination of some of Ketcheson's data with changes in population sentiment. KAUST Associate Professor Xiangliang Zhang and her team are collecting data from millions of social media posts to understand trends in social sentiment throughout the pandemic. Their data—alongside that of colleagues and collaborators from around the world—could help inform public health interventions, institutional responses and more.Other KAUST faculty involved include Professors Carlos Duarte, Arnab Pain, Hernando Ombao and Xin Gao and newly recruited faculty member Paula Moraga, a current lecturer at the University of Bath in the U.K.