Achievement

Sensitivity analysis quantifies the effects of parameter perturbations on system behavior, which can be used to identify important reactions in complex signaling networks. Sensitivity analysis in stochastic models of biological systems is limited due to the high computational costs needed to perform many Monte Carlo simulations. IGERT trainee Patrick Sheppard developed two new efficient stochastic sensitivity methods that significantly reduce these costs: the common reaction path method for finite parameter perturbations, and the regularized pathwise derivative method for infinitesimal perturbations. Both methods exploit the random time change representation of stochastic models to reduce variance of Monte Carlo estimates. He developed numerical algorithms to enable implementation of the methods, demonstrating speedups of several orders of magnitude have been demonstrated for some problems.