Abstract (eng)
The evaluation of protein crystal structures is crucial for understanding protein function and drug development. The analysis of single crystal structures is well established, especially for small chemical compounds. In contrast, visualising and aggregating the relevant information of
the variety of whole protein ensembles is still rare and time consuming to perform manually.
Since a plethora of protein-ligand complex structures are available in pharmaceutical industry, the analysis of protein crystal structure ensembles has become a new task in the field of drug
design. Furthermore, frequent mutations in the binding site of oncogenic targets additionally enhances the requirement for fast analysis tools.
In this study, a tool kit to rapidly access and exploit the available 3D structural knowledge was developed. The visualisation and qualitative analysis provides an overview of structural differences and should help to drive decisions for drug design. Grid maps are the main topic of this thesis. They are defined by values of certain structural properties distributed on a 3D grid. Thus, properties can be stored and manipulated, such as potential binding energies of protein-ligand complexes. Among the possible applications of grids in medical chemistry, protein-ligand interaction ”hot spots” are relevant for rational design of new compounds. Hot spots are calculated potential energies, indicating high propensity of a target protein for compound binding, and can be represented as contour surfaces. These interaction hot spots define favourable regions
for binding of certain compound features and are crucial to form protein-ligand complexes.
The tool kit presented in this thesis provides insight into interaction hot spots of protein-ligand complex ensembles and thus it offers new compound design opportunities. The algorithm is able to extract single hot spots according to a new crystal structure as reference, enabling
a spatial resolution of relevant hot spots. These results can be compared pairwise with an ensemble applying the grid analysis script. In addition, it provides combined information of visual and qualitative analysis of 3D grid maps. A heat map representation of the calculated energy differences and similarities assists to evaluate the knowledge gain.