Even though it has been entirely eradicated from North America, malaria is still a major cause of death in developing countries throughout Africa, Southern Asia, and South America. These regions do not have sufficient access to the vaccines widely available to richer areas. In fact, 3.3 billion people worldwide live in areas at risk of malaria transmission . Distributing already-available forms of treatment is simply too costly to reach every individual at risk of ac-quiring malaria. The continued survival of this disease has sparked a debate in the ecological community: what is the most effective way to perpetually halt the spread of malaria ?
Multiple plans have been proposed, including ideas of eradicating entire mosquito populations, solely targeting mosquito larvae, and even using a “mosquito vaccine” that prevent mosquitoes from being able to carry plasmodia, the parasitic protists that cause malaria [2, 3]. How-ever, progress in experimenting with these treatment plans has been difficult. Such massive pro-jects require extraordinary investments in both time and money. Furthermore, thousands of lurk-ing variables, including everything from immunity to malaria in parts of the human population due to the presence of a sickle cell allele to patterns of rainfall across the region in question make these experiments impossible to control. One of the only options left in determining the best method to eradicate malaria is to rely on the digital world over that of the physical.
In the past two decades, computers have been used for everything from managing bank accounts to communicating via social media. Science has also witnessed a similar “rise of the machine,” as increased computational power advanced numerous disciplines. Both the speed and efficiency of computational power have expanded the use of computer modeling that applies theoretical data and mathematical equations. These models provide an alternative to physical experimentation, a process that can be both expensive and time-consuming .
Computer models can be applied to a multitude of purposes, with a range of simulations each suiting a specific situation. For example, agent-based models, also known as ABMs, are a specific form of computational model. Developed in the early 1940s, ABMs take into account the actions of autonomous agents to determine the progression of the whole system. They are particularly useful due to their ability to simulate the heterogeneity of the actors in the model: agents of different types do not necessarily have to act at the same time as each other. The basic interac-tions between individuals combine to reveal a collectively chaotic behavior. ABMs have tremendous application to a variety of real-world issues due to their ability to model what would other-wise be complex systems.
Their deterministic behavior closely matches that of many real-world scenarios, such as those in the social sciences. For instance, in 1971, Thomas Schelling created an agent-based model that reproduced racial segregation in households in American cities. Similar uses for agent-based models in the social sciences include opinion dynamics, industrial networks, and supply chain management . ABMs have been heavily applied to everything from meteorology to politics. However, it appears that ABMs may have the greatest impact on society through their applicability to ecology and the environmental sciences. Particularly, agent-based models may solve the problem mentioned earlier: eradication of malaria.
Malaria is a type of zoonotic disease, an illness caused by an infectious agent that is transmitted across species. The pathogen in this situation is plasmodium, a parasitic protozoa that vector organisms, like mosquitos, can transfer to hosts, including humans. This behavior follows that of the “predator/prey” class of agent-based models, where certain individuals prey upon or infect other individuals. In modeling the spread of malaria through agent-based models, mosquito agents and human agents would populate the space being simulated. Certain behavior would drive each class of individuals, and activities including reproduction and death would be simulat-ed. A combination of environmental and agent-based variables, whose values would be driven by related data, would be inputted into the model.
Of course, determining the exact formulae that sets the values such as maturation rate of mosquito larvae or the probability of infection given that an individual has been bitten by an in-fectious mosquito becomes fairly subjective. This requires extensive discussion and exploration in the scientific community. With the vast amount of data available online today, it should be easy to examine the trend in values. Yet, deciding upon the precise equation that would best fit the trend is essentially up to persistent trial and error.
In determining the equations that influence the agents in the model, a combination of en-vironmental factors, including geography, humidity, and immunity of populations, and experi-mental variables, such as the type of treatment applied, would be needed. A variety of treatment plans could be used, including the introduction of genetically-modified mosquitoes that are im-mune to plasmodia or radiating mosquito larvae to turn them sterile in adulthood. Regardless of what treatment is modeled, the advantage of simulating such an experiment is that control of variables is established, preventing lurking variables from affecting results. Ultimately, while any form of computer simulation is key to making progress in determining the right control strategy for malaria, it is clear that the agent-based model has huge potential to revolutionize the search for the best eradication technique for the disease.
References “Malaria Facts.” Centers for Disease Control and Prevention. Last modified September 19, 2012.  Fang, Janet. “Ecology: A world without mosquitoes.” Nature, July 21, 2010, 432-34.  Szalavitz, Maia. “Hopes for a New Kind of Malaria Vaccine.” Time, January 15, 2010.  Codling, Edward A., and Alex J. Dumbrell. “Mathematical and theoretical ecology: linking models with ecological processes.” Interface Focus 2, no. 2 (April 2012): 144-49.  Gilbert, Nigel. “The Idea of Agent-Based Modeling.” In Agent-Based Models, 1-20. N.p.: n.p., 2008.  Hay, Simon I., and Robert W. Snow. “The Malaria Atlas Project: Developing Global Maps of Malaria Risk.” PLoS Med 3, no. 12 (December 5, 2006).  Kennedy, Ryan C., Kelly E. Lane, S.M. Niaz Arifin, Agustín Fuentes, Hope Hollocher, and Gregory R. Madey. “A GIS Aware Agent-Based Model of Pathogen Transmission.” In-ternational Journal of Intelligent Control and Systems 24, no. 1 (March 2009): 51-61.
Image Credit: Retrieved February 23, 2014 from: Wikimedia Commons. Labeled for reuse.