Qualitative and semi-quantitative analysis of signal transduction networks

In this work, new methods for qualitative modeling and structural analysis of signaling networks and for data-driven network interrogation are presented. Interaction graphs and static logical models are considered as qualitative modeling frameworks. With applications to two central signaling pathways for liver regeneration, it is shown how these methods can contribute to unravel structure and functioning of cellular signaling systems. Besides purely qualitative modeling approaches, the potential of combining qualitative and quantitative mathematical is shown by presenting a new hybrid modeling approach which combines interaction graph and ODE modeling. Furthermore, new algebraic and graph-theoretic criteria that enable to exclude certain dynamic behavior of variables in an ODE model independently of concrete parameter values are presented. In particular, criteria for the class of Dynamic Chemical Reaction Networks were derived. These methods can be used to preselect model structures. The work demonstrates the potential of qualitative modeling for the analysis of signaling networks. In particular, it is demonstrated how different modeling approaches complement each other, and how results obtained by qualitative modeling methods can serve as a base for more detailed quantitative modeling.