BioUML plug-in for nonlinear parameter estimation using multiple experimental data

Motivation: Systems biology deals with many different types of experimental data representing individual components of biological systems. Behavior of these systems over time could be described using systems of ordinary differential equations (ODE). In order to analyze dynamics of the ODEs and estimate their parameters based on data obtained in different experimental conditions, biologists need a flexible framework that allows them to create dynamic models and perform multi-experiment parameter fitting. Results: We present optimization tools of the BioUML software ( http://biouml.org ) developed for modeling and analysis of biochemical systems. We created optimization plug-in to solve non-linear optimization problems via minimization of the function of deviations between experimental data and model simulation results. Experimental data can be considered as separate sets of time courses or steady states stored in different tab-separated files. BioUML includes several deterministic and stochastic optimization methods which find reasonably accurate solutions faster than the COPASI software. Some of these methods provide constrained optimization and some of them were parallelized.