Stochastic modelling provides an indispensable tool for understanding howrandom events at the molecular level influence cellular functions. In practice,the common challenge is to calibrate a large number of model parameters againstthe experimental data. A related problem is to efficiently study how thebehaviour of a stochastic model depends on its parameters, i.e. whether achange in model parameters can lead to a significant qualitative change inmodel behaviour (bifurcation). In this paper, tensor-structured parametricanalysis (TPA) is presented. It is based on recently proposed low-parametrictensor-structured representations of classical matrices and vectors. Thisapproach enables simultaneous computation of the model properties for allparameter values within a parameter space. This methodology is exemplified tostudy the parameter estimation, robustness, sensitivity and bifurcationstructure in stochastic models of biochemical networks. The TPA has beenimplemented in Matlab and the codes are available at http://www.stobifan.org .
展开▼
