Applying a global sensitivity analysis workflow to improve computational efficiencies in physiologically-based pharmacokinetic model


Traditionally, the usual solution to reduce parameter dimensionality in the physiologically based pharmacokinetic (PBPK) model is through expert judgment. However, this approach may lead to lower efficiency and substantial bias in parameter estimates. The purpose of this study is to propose a global sensitivity analysis (GSA) algorithm to ascertain which parameters in a PBPK model are non-identifiable, and therefore can be assigned fixed values to improved speed and convergence in Bayesian parameter estimation with minimal bias. Overall, we propose that this GSA approach provides an objective, transparent, and reproducible approach to improve the performance and computational efficiency of PBPK models.

57th SOT Annual Meeting