Inference method for a stochastic target-mediated drug disposition model via ABC-MCMC

Abstract: Inference method for a stochastic target-mediated drug disposition model via ABC-MCMC In this study, we discuss model robustness. Model robustness is consistent performance over variations of parameters. We formulate a stochastic target-mediated drug (TMDD) model, one of the pharmacokinetic models, to capture bi-exponential drug decay in plasma. A stochastic process is used to account for system randomness, and this process is transformed into system of stochastic differential equations. Parameter inference is performed by Approximation Bayesian Computation using the likelihood-free method. Using these collected samples, global sensitivity of parameters is compared to Uniform and Normal distributions. This approach in the TMDD model may improve model robustness without changing the global sensitivity of parameters and the model.