MODEXA

In dem interdisziplinären Verbundforschungsprojekt MODEXA sollen modell­gestützte Metho­den und Werkzeuge zur optimalen Planung von Experimenten mit dem Ziel der quantitativen Aufklärung der Struktur und Dynamik von Signal­trans­duktions­­­kaskaden entwickelt werden. Als biomedizinisch relevantes System wird die Signaltransduktion der unter genotoxischer Belastung induzierten NF-kappa B Regulation in Säugerzellen betrachtet. Es ist beabsichtigt, innovative experimentelle Techniken und neue systemtheoretische Methoden zu entwickeln, um gültige quantitative Modelle zur Beschreibung der NF-kappa B Signaltransduktion bei genotoxischer Belastung zu generieren. Unter Nutzung standardisierter Schnittstellen soll ein modular strukturiertes Software-Werkzeug, welches mathematische, informations- und anwendungstechnische Aspekte gleichermaßen berücksichtigt, entwickelt und anschließend kommerziell als MODEXA-toolbox vermarktet werden.

Model-based Methods for the optimal Design of Stimulus Experiments and Dynamic Analysis of Signal Transduction Processes

The interdisciplinary research project MODEXA aims at developing model-based methods and tools for the optimal design of experiments on living cells with the goal of quantitative elucidation of structure and dynamics of signal transduction cascades.
Stimulus-experiments on cells, development of network models for the signal transduction networks under investigation and the model-based planning of experiments are linked in an iterative workflow, enabling the identification and parameterization of adequate network models.
As an example of high biomedical relevance the genotoxic stress-induced NF-kappa B regulation in mammalian cells is considered. The dynamic response behaviour of the cells is determined via target molecules, which can be tracked quantitatively over time in the cytosol as well as in the nucleus. This allows the formulation of compartment models, which describe the cellular signal transduction cascade by means of coupled systems of ordinary differential equations (ODEs).
In the course of model development competing network alternatives with different levels of complexity will be generated, from which the adequate model has to be selected step-by-step (model discrimination). In addition, the kinetic parameters of the formulated models have to be estimated with highest possible precision (parameter identification). Both tasks can be accomplished by designing the experimental stimuli in such a way that the generated system responses contain the maximum of information for either purpose. To this end gradient-response methods (GMs), non-linear frequency response analysis (FRA) as wells as the sigma-point method (SPM) are introduced into systems biology as innovative techniques for optimal experimental design.
Experimental data structures, the generation of model alternatives, model discrimination and parameter identification are intended for implementation into a modular software tool using established interfaces.