Approximate Bayesian computation for the Ross-Macdonald model of malaria transmission

Abstract

The Ross-Macdonald model remains a key component in understanding the dynamics of malaria transmission, representing the complex interactions between mosquito and human populations through a coupled system of ordinary differential equations (ODEs). Parameter estimation for this model presents significant challenges due to the intractable likelihood function and multiple uncertain parameters that govern transmission dynamics in epidemiology. This study implements Approximate Bayesian Computation with Sequential Monte Carlo (ABC-SMC) methods to perform parameter inference on the Ross-Macdonald model. We demonstrate that ABC-SMC provides a robust framework for parameter estimation while accounting for model uncertainty and parameter correlations. Our analysis reveals critical information on the sensitivity of the basic reproduction number R₀ to various epidemiological parameters. The methodology presented offers a practical approach to the development of evidence-based policies and provides uncertainty quantification essential for robust decision making.