SBMLSimulator: A Java Tool for Model Simulation and Parameter Estimation in Systems Biology.

Andreas Dräger,Roland Keller,Andreas Zell,Alexander Dörr

doi:10.3390/computation2040246

Abstract

The identification of suitable model parameters for biochemical reactions has been recognized as a quite difficult endeavor. Parameter values from literature or experiments can often not directly be combined in complex reaction systems. Nature-inspired optimization techniques can find appropriate sets of parameters that calibrate a model to experimentally obtained time series data. We present SBMLsimulator, a tool that combines the Systems Biology Simulation Core Library for dynamic simulation of biochemical models with the heuristic optimization framework EvA2. SBMLsimulator provides an intuitive graphical user interface with various options as well as a fully-featured command-line interface for large-scale and script-based model simulation and calibration. In a parameter estimation study based on a published model and artificial data we demonstrate the capability of SBMLsimulator to identify parameters. SBMLsimulator is useful for both, the interactive simulation and exploration of the parameter space and for the large-scale model calibration and estimation of uncertain parameter values.

Highlights

The dynamic simulation of quantitative biological models belongs to the key aspects of research in systems biology [1]
Such parameter estimations will often be conducted in the command-line mode, which is an alternative to the graphical user interface (GUI) that SBMLsimulator launches by default
The graphical user interface comprises several separated sub-windows for the presentation of simulation results as well as for settings that can be modified by the user: at the lower part of the GUI, SBMLsimulator enables the user to choose the numerical solver and to set the start and end point as well as the step size of the simulation

Summary

Introduction

The dynamic simulation of quantitative biological models belongs to the key aspects of research in systems biology [1]. A graphical display of the resulting curves can greatly facilitate the analysis of the system The calculation of these dynamics requires the initial values of each model component to be known, including the concentration of reactive species as well as parameters, such as Michaelis constants. Expression Omnibus [9] or KiMoSys databases [10]) To this end, heuristic optimization routines can be applied to fit the model to the experimental data. Many systems biology simulation and optimization frameworks are available, e. The focus of SBMLsimulator is to provide the scientific community with an usable and flexible parameter estimation tool that understands and supports all aspects of the modeling format SBML through all of its levels and versions. Teams of core developers maintain both open-source libraries SBSCL and EvA2 that can be extended by the scientific community

Implementation

Results and Discussion

Graphical User Interface

Parameter Estimation Study

Conclusions

Availability and Requirements