Multi-method global sensitivity analysis of mathematical models

Authors: Do Dela A
Publication: ScienceDirect
Keywords: Derivative-based global sensitivity measures, eFAST, Global parameter sensitivity analysis, HIV model, Sobol’s method, Tumor growth model
Division: Quantitative Systems Pharmacology (QSP)

Highlights

  • Selecting appropriate parameter sensitivity methods poses significant challenges.
  • We develop a framework that employs and compares three different methods.
  • We provide an intuitive understanding of each method via MATLAB codes.
  • A guide for tuning algorithm parameters is created and presented.
  • New methodology demonstrated on models of HIV progression and tumor growth.