For a mathematician like Aaron King, nothing beats an algorithm for understanding how the world works. Especially the natural world. King is part numbers guy, part ecologist, and part of a new generation of ecologists using algorithms to describe the incredibly complex relationships that develop over time among the component parts of an ecosystem. For King, this means humans and diseases, specifically.

“In ecology, we come to grips with fundamentally complex systems,” King says. “The only way we, as human beings, can make progress with that kind of complexity is to think things through carefully, logically, and mathematically.”

The mathematically challenged may find it hard to believe, but King says developing the algorithm is the easy part of his work. The hard part is testing the algorithm against real-world data to fill in all the unknown variables. King methodically tracks down data on hospitalizations, vaccinations, and deaths associated with a particular disease in order to study how disease pathogens infect and interact with their human hosts.

“It’s a good thing human beings have this tendency to write everything down,” says King, an assistant professor of ecology and evolutionary biology and of mathematics, who is also affiliated with LSA’s Center for the Study of Complex Systems. “I thank God for bureaucracy every day, because bureaucrats keep records,” King adds. “These long-term data sets are the reason we’ve been able to make progress.”

King joined the LSA faculty in 2005 and soon met Edward Ionides, an associate professor of statistics. The two researchers began working together to develop a set of algorithmic tools that model how ecosystems work. King posts these algorithms on the Internet, where they are available for use by scientists and statisticians around the world.

“The algorithms are broadly applicable,” King says. “They can be used in finance, fisheries management, baseball—anything where you have dynamics that play out through time and some sort of time-series data.”

King has studied diseases such as measles and cholera, the latter in collaboration with Mercedes Pascual, professor of ecology and evolutionary biology, and others. But his current research with Pej Rohani, professor of ecology and evolutionary biology, concentrates on a major public health mystery: Why are cases of pertussis, or whooping cough, increasing in the United States and other developed countries, in spite of widespread use of preventive vaccines?

Mothers in the early 1900s knew all too well what it meant to hear their babies coughing in the night. Pertussis was a common childhood disease, and it was often lethal, especially to infants and young children. Click here to hear the whooping sound of a baby with pertussis.

After pertussis vaccines were developed in the 1940s, the number of cases dropped dramatically. But in the 1980s, public health experts began noticing something odd: Cases of whooping cough were increasing, especially in previously vaccinated adolescents and young adults. And no one knew why.

King’s research is more than just an academic exercise. It could help public health experts target booster vaccination programs to the people most likely to become infected and spread the disease to others—especially babies, who are most at risk from whooping cough.

Researchers have lots of ideas on possible reasons for the pertussis comeback. PerhapsBordetella pertussis, the bacterium that causes the disease, has acquired genetic mutations that make it more infectious. Maybe the vaccine loses its effectiveness over time. The problem, according to King, was that none of the theories explained why pertussis seemed to strike certain age groups, but not others.

“One key question that’s still unresolved is: Does vaccination protect you from being infected or does it just protect you from the disease itself?” In other words, “You can still become infected and transmit the disease, but you won’t get sick,” King says.

In an attempt to answer that question, King and his collaborators developed two models and tested them against two data sets, one from Sweden and one from Massachusetts. Unfortunately, the results turned out to support two opposing conclusions, but King doesn’t appear to be discouraged.

“We have two models and they cannot both be right,” he says. “Probably both of them are wrong, but maybe interestingly wrong. It’s a process and we keep moving toward the goal.”