It is still dark at 7am, as I run along the beach in the southernmost region of Cape Town, South Africa. The winter air is crisp and sharp, and the frosty waves that lap at my tennis shoes roll off to distant Antarctica. I am breathless from the run and the chill and my awe of the vast world around me.
Putting Meaning Into Modeling
My colleague from the University of Antananarivo, Christian Ranaivoson, and I have taken a brief respite from our doctoral research tracking zoonotic bat viruses in Madagascar. In lieu of countless hours of bat netting in remote corners of the Malagasy rainforest, we find ourselves immersed instead in mathematics.
We’re attending the sixth annual Clinic on the Meaningful Modeling of Epidemiological Data (MMED). Its goal is to bring together mathematicians who can work wonders processing data and biologists who study disease outbreaks and can interpret that data as it applies to the real world.
Case in Point
On day one of the workshop, our professors present us with a series of graphs showing the number of cases of various diseases throughout time in London and Wales. We are asked to talk to our neighbors about what we think is going on in the images.
“Well, obviously hygiene and antibiotics have greater impacts on the bacterial pathogens,” I say gesturing to the typhoid and scarlet fever graphs. “But sheer number of susceptible hosts is more important for the virus—you can see the spike in measles cases during the baby boom. More babies, so more infections.”
“Sorry, but which graph shows a virus?” asks my classmate, a South African mathematician. “And what is the baby boom?”
History of western civilization aside, my classmate has been trained as a computer programmer and is likely far more capable in the quantitative analysis of infectious disease than I am. But the MMED program exists to remind us that science is an interdisciplinary endeavor and requires crosstalk among many fields in order to make progress. Drawing students from across Africa plus a few North American attendees, MMED strives to teach scientists who collect data about the power of mathematical modeling and teach the modelers about the meaning behind the data. “You can always find someone to do the math,” says MMED professor and former WHO researcher, Brian Williams. “But to make the connections and advance the ideas—that is something truly unique.”
The Stories Behind the Data
Just last month, I heard a talk about specific kinds of mice (all in the species Mastomys natalensis, the multi-mammate mouse) which each host a different species of arenavirus. One subgroup of these mice in West Africa serves as the natural reservoir host for the Lassa virus, which causes hemorrhagic fever and can pass to humans. The disease seems to be restricted to West Africa because that’s the only place this specific genotype of the mouse species lives–though other genotypes range broadly across the continent. It turns out that something seemingly unrelated, like the taxonomy of a mouse, matters for understanding the spread of a virus. In biology it’s all connected intimately. In the words of my favorite author, John Steinbeck, “None of it is important or all of it is.”
At one of our interminable tea breaks (a holdover from British colonialism), I listen to Brian explain that new cases of disease-state tuberculosis take place disproportionately in individuals with low body mass index (BMI). “There’s a reason TB is called consumption,” Brian says; the disease quite literally eats a person up. “But still, we don’t really understand why this association is specific to TB,” he continues, and I find myself joining in the debate.
Malnutrition is known to compromise our cells’ immunity functions, and TB is a bacterial pathogen that works inside cells—so it makes sense that TB might have more noticeable disease symptoms in malnourished individuals. But the interaction is complicated because the bacterium also requires lipid resources inside the cell to survive.
“TB eats intracellular lipids, just as the protozoan that causes malaria eats hemoglobin in the red blood cell,” I venture.
Thinking one step further, I wonder whether this might be the same with other intracellular pathogens, too. Most of them are viruses, however, and scientists still debate whether or not they are even alive, much less what they eat. Viruses need intracellular space to multiply, but their resource requirements are very different from those of bacteria or protozoa.
“So we shouldn’t expect the BMI association to necessarily repeat itself with other pathogens,” I say.
“That’s brilliant,” says Brian.
“That’s ecology,” I laugh. Thinking of the whole environment and the way different players interact in it is key to understanding biology, but not necessarily the first thing a mathematician might think of. As in the case of the infection time lines above, we learn that the model must be adapted to the biology.
Bringing It All Home
As a Princeton PhD student, I feel constantly torn between theory and application; I find myself sad to depart from the intellectual stimulation of MMED, reluctant to leave behind the long hours of coding, data analysis, and paper reading that made up my life of the past two weeks. And yet, I have learned that with modeling there must also be meaning, and for me, that meaning is clear as day on the Eighth Continent.
“Princeton will give you a solid theoretical foundation,” Jonathan had told me as the clinic drew to a close. “But it is those who dare to take that foundation and grapple with the real world that make a difference.”
These words are fresh in my mind as Christian and I board our plane back to Madagascar where one more month of mud, mosquitoes, and bat netting awaits us. We return to a land of airline strikes and delayed flights, of back-ordered reagents and broken liquid nitrogen tanks, and once again, I find myself reworking our field plans on the fly. But now, I have justification for my decision to work in such a challenging environment: I’m taking my theories and grappling with the real world in the truest sense there is.
As we walk across the tarmac at Ivato International Airport, once more home-sweet-home in Madagascar, I grin at Christian and ask, “Remember that time we went to South Africa for a math class?”
“Yes,” he nods back. “But it was much more than that.”
The Clinic on the Meaningful Modeling of Epidemiological Data is run jointly by the University of Florida, the South African Centre for Epidemiological Modelling and Analysis (SACEMA) and the African Insitute for Mathematical Sciences (AIMS) in Muizenberg, Cape Town, South Africa.
Thanks to the MMED faculty: Steve Bellan, Faikah Bruce, Jonathan Dushoff, John Hargrove, Wilfred Ndifon, Travis Porco, Juliet Pulliam, Jim Scott, Alex Welte, and Brian Williams. Learn more at: http://www.ici3d.org/mmed/