OpenMalaria and EMOD: a case study on model alignment
Casey Ferris, Benoit Raybaud, Gregory R. Madey · 2015
This paper presents a case study of model alignment involving two independently developed agent-based malaria models. Each of these models simulates malaria transmission and various malaria interventions, such as insecticides, drugs, and bednets. EMOD specifically models both humans and mosquitoes as agents that interact with each other and interventions. However, OpenMalaria, only directly models humans as agents, while the mosquitoes interact based on results from several differential equations. This is the main difference between the two models, but there are several more, some which are not clearly noticeable without performing a complete comparison. In this docking exercise, we take a unique third-party perspective in order to determine equivalence between these two models. The successful alignment of OpenMalaria and EMOD further validates each model beyond their previous, individual validations. This case study also provides insight into the challenges and benefits of docking when none of the original model creators are fully involved in the exercise.