Using Multicriteria Analysis of Simulation Models to Understand Complex Biological Systems

Scientists frequently use computer-simulation models to help solve complex biological problems. Typically, such models are highly integrated, they produce multiple outputs, and standard methods of model analysis are ill suited for evaluating them. We show how multicriteria optimization with Pareto optimality allows for model outputs to be compared to multiple system components simultaneously and improves three areas in which models are used for biological problems. In the study of optimal biological structures, Pareto optimality allows for the identification of multiple solutions possible for organism survival and reproduction, which thereby explains variability in optimal behavior. For model assessment, multicriteria optimization helps to illuminate and describe model deficiencies and uncertainties in model structure. In environmental management and decisionmaking, Pareto optimality enables a description of the trade-offs among multiple conflicting criteria considered in environmental management, which facilitates better-informed decisionmaking.