The analysis of the structural and the dynamical behavior of biomolecules is very important to under- stand their biological function, stability or physico-chemical properties. In this thesis, it is highlighted how different theoretical methods to characterize the aforementioned structural and dynamical properties can be used and combined, to obtain kinetic information or to detect biomolecule-ligand interactions. The basis for most of the analyses, performed in the course of this work, are molecular dynamics sim- ulations sampling the conformational space of the biomolecule of interest. Using molecular dynamics simulations, the remarkable stable water-soluble-binding-protein is examined first. On a theoretical ba- sis, structural modifications that can influence the stability of the protein are discussed. Additionally, by combining the simulations with a QM/MM optimization scheme and quantum chemical calculations, spectroscopical properties can be investigated. Markov State Models are applied frequently to capture the slow dynamics within simulation trajectories. They are based on a discretization of the conformational space. This discretization, however, introduces an error in the outcome of the analysis. The application of a core-set discretization can reduce this error. In this thesis, it is discussed how density-based cluster algorithms can be used to determine these core sets, and the application on linear and cyclic peptides is highlighted. The performance of a promising cluster algorithm is investigated and error sources in the construction of the Markov models are discussed. Finally, it is shown how molecular docking combined with molecular dynamics simulations can be used to determine the binding behavior of ligands towards biomolecules. In this context, the important in- teractions within the active site of an enzyme, and different binding modes of DNA intercalators are identified.
