KinFitSim® 3.0: Kinetics simulation and fitting software

Mathematical Methods

KinFitSim 3.0 uses a version of the Gear’s method as a basis for its automatic module for the solution of ODEs that describe time evolution of concentrations of species involved in a homogeneous (bio)chemical process. It automatically detects whether the ODE system is stiff and solves it using adjustable time steps and variable order predictor-corrector algorithm. Thus in most practical situations the user does not need to set any parameters regarding the numerical simulation.

The manual simulation regime may be used in the unlikely event of failure of the automatic simulation procedure or for learning the effects of changing the numerical method and its parameters on the obtained results. In this case there is a choice between the variable step-size 6th order Runge-Kutta method and Gear’s method and a possibility to control the method’s relative accuracy.

The fitting part of the program provides a choice among four available minimization methods, i.e. the Nelder-Mead, steepest descent (modified gradient search), Marquardt and Gauss-Newton methods. All methods are well suited for the minimization of least-squares functionals that reflect the difference between simulated and experimental data to be minimized. The default fitting options (that include the Nelder-Mead method) allow handling the majority of problems without the need of changing method or its parameters.