SP10 - Network Modeling

T. Beissbarth, C. Bender, A. Jöcker (DKFZ)

Our main objective is to model signaling pathways in the context of cancer cells, based on qualitative and quantitative data collected and analyzed within IG-CSG.

• The generation and extension of models from literature, screening and interaction data, using machine learning approaches and Boolean logic models.
• The development of a full Boolean description of cancer related signaling related to cell proliferation
• The analysis and simulation of models
• The development of methods for the quantitative analysis and pre-processing of experimental data from protein arrays

These objectives contribute directly to the development of a comprehensive, quantitative and dynamic mathematical model of cancer-related signal transduction processes in cell cycle regulation, which shall serve as the basis for the identification of new therapeutic strategies.

We have developed a new method for the extension of models of signal transduction methods by predicting membership of individual proteins to pathways (Fröhlich 2008, Hahne 2008). A Boolean model of cell cycle control in breast cancer has been derived using network reconstruction approaches, and the derived model has been compared with a network derived from literature. Based on this model, potential targets for therapeutic intervention have been proposed (Sahin 2009).

Visualization of the time course of Erk1/2 phosphorylation upon stimulation of cells with EGF (t=0), after pre-incubation of cells with different drugs alone and in combination.

Next we will extend the Boolean network of ERBB related signaling (mostly with SPs 3,4,7) and then translate this into quantitative and dynamic models using differential equations, integrating both literature knowledge and the models derived in the first funding period using machine learning methods.