Intracellular signal transduction occurs in very complex biochemical reaction networks and often cannot be comprehended intuitively. The mathematical modelling of such networks via ordinary differential equations allows their investigation with the aid of systems theoretical methods. The first main part of this thesis handles the mathematical modelling of the NF-?B signal transduction pathway, and its analysis in terms of the biological importance of certain mechanisms. The second main part of the work deals with methodological challenges in the mathematical modelling of signal transduction networks and introduces a modelling approach which could contribute to the solution of these problems.In this thesis, the TKM approach is extended to be used with reaction rules, developing a method which combines the rule-based modelling of reaction networks with thermodynamical consistency and formalised model reduction. Interaction factors are a crucial concept within this approach, as they allow parameterising the impact of bindings on the equilibrium and the velocity of reactions. By using this concept, the number of model parameters that need to be defined can usually be drastically reduced, while the flexibility of parameterisation is considerably enhanced. The effects of symmetric complexes on the parameterisation are also taken into consideration. An implementation of the approach in Mathematica is briefly described. Furthermore, an extension of the approach is outlined which, by reduction at the rule level, further improves the efficiency of the algorithm. The approach is illustrated using a model of phosducin phosphorylation in the light adaptation of the eye.