A New Method for Ligand-supported Homology Modelling of Protein Binding Sites: Development and Application to the neurokinin-1 receptor
In this thesis, a novel strategy (MOBILE (Modelling Binding Sites Including Ligand Information Explicitly)) was developed that models protein binding-sites simultaneously considering information about the binding mode of bioactive ligands during the homology modelling process. A...
Keine Tags, Fügen Sie den ersten Tag hinzu!
|Zusammenfassung:||In this thesis, a novel strategy (MOBILE (Modelling Binding Sites Including Ligand Information Explicitly)) was developed that models protein binding-sites simultaneously considering information about the binding mode of bioactive ligands during the homology modelling process. As a result, protein binding-site models of higher accuracy and relevance can be generated. Starting with the (crystal) structure of one or more template proteins, in the first step several preliminary homology models of the target protein are generated using the homology modelling program MODELLER. Ligands are then placed into these preliminary models using different strategies depending on the amount of experimental information about the binding mode of the ligands. (1.) If a ligand is known to bind to the target protein and the crystal structure of the protein-ligand complex with the related template protein is available, it can be assumed that the ligand binding modes are similar in the target and template protein. Accordingly, ligands are then transferred among these structures keeping their orientation as a restraint for the subsequent modelling process. (2.) If no complex crystal structure with the template is available, the ligand(s) can be placed into the template protein structure by docking, and the resulting orientation can then be used to restrain the following protein modelling process. Alternatively, (3.) in cases where knowledge about the binding mode cannot be inferred by the template protein, ligand docking is performed into an ensemble of homology models. The ligands are placed into a crude binding-site representation via docking into averaged property fields derived from knowledge-based potentials. Once the ligands are placed, a new set of homology models is generated. However, in this step, ligand information is considered as additional restraint in terms of the knowledge-based DrugScore protein-ligand atom pair potentials. Consulting a large ensemble of produced models exhibiting di erent side-chain rotamers for the binding-site residues, a composite picture is assembled considering the individually best scored rotamers with respect to the ligand. After a local force-field optimisation, the obtained binding-site models can be used for structure-based drug design.|