Computational Studies on the Relation Between Macromolecular Dynamics and Protein Binding and Function

Computational methods can help to better understand and analyze the interaction of proteins and their binding partners. This interaction is influenced by many factors, including specific sequence variants, the dynamics and electrostatics of the proteins, as well as further physicochemical properties of the corresponding binding partners. A detailed investigation of these different, and often complicated, properties helps to better understand the functionality of proteins, for which the interaction with other molecules plays a crucial role. The work presented here provides new methodologies, implemented in webservers and software, which assist during the analysis of proteins. Furthermore, in an application case, computational methods and analyses in combination with experimental results were used to detect a specific interaction network of proteins. The new ProSAT+ webserver enables the visualization of protein sequence annotations in the context of the three–dimensional protein structure and contains additional options for visualizing and sharing protein annotations. The sequence information allows an easy, but extensive analysis of proteins. The functionality of the ProSAT+ webserver can be integrated into other webservers, which was done in the case of the two other webservers for the analysis of protein binding pockets described here. A tool for the LigDig webserver was developed that provides the comparison of protein binding pockets by the alignment and visualization of the binding pockets based on an existing algorithm. The new TRAPP webserver assists in the analysis of protein binding pocket dynamics. The existing TRAPP software was used, and a user web interface was implemented to simplify the usability. Additional new functionalities were also developed, such as the visualization of protein sequence conservation in context of all other TRAPP results in the three–dimensional structure. This allows the detection of conserved or non–conserved regions inside the binding pocket, which might influence the dynamics of the pocket. This newly gained information can be used during the process of designing selective inhibitors. During the protein disaggregation process, members from different classes of the so-called J–protein (HSP40) co–chaperones play a crucial role. The synergetic application of different computational methods and experiments enabled the detection of an interclass specific J–protein interaction and indicated that the interaction evolved to enable a high efficiency in the disaggregation process. The resulting data of performed protein domain docking simulations required an update of the standard clustering workflow. This new methodology can be applied for protein docking in cases that have problems with multiple, weakly specific interaction sites. The work presented here facilitates in many ways the analysis of proteins, including their structure and sequence features, as well as, their dynamics and interactions with their binding partners. The new methods are provided as webservers and therefore are accessible, and easy to use for all researchers. This can assist in many research projects and provide relevant information. The analyses of the J–proteins improved the knowledge about their biological role and functionality, and therefore provide an important contribution for a better understanding of the overall protein disaggregation process