Hydrogen bonds (H-bonds) are a mayor driving force of protein-ligand recognition. A profound understanding of these interactions – why they form and which geometrical preferences they have – is of interest in various life science applications. Analyzing statistically relevant data, quantitatively as well as qualitatively, leads to chemical meaningful conclusions with implications for medicinal chemistry. Here, we present a new study on a large experimental data collection. The data set contains about 9,000 high resolution protein-ligand complexes extracted from the Brookhaven Protein Data Bank (PDB). Using this data set, we thoroughly investigated the geometric and functional properties of H-bonds and their dependence on the chemical environment. We defined an extensive range of functional groups relevant for protein and ligand interactions and included diverse structural and chemical properties of their surroundings. Accompanying the statistical analysis was the development of the software NAOMInova, which allows non-computational scientists to search for and browse through interaction geometries. NAOMInova allows the definition of a functional group of interest and presents the distribution of potentially interacting partner atoms found in crystal structures. Additionally, NAOMInova provides a backlink to the underlying structure enabling an easy retrieval and visualization of the original protein-ligand complex. This way, outliers as well as further structural properties can be analyzed and conclusion for ligand optimization can be derived. In this talk, we will present results from our statistical study, including unexpected geometric patterns and newly derived optimal H-bonds interaction geometries. We show that the application of NAOMInova intuitively aids the identification of common and untypical interaction geometries. In this way, the medicinal chemist is supported in understanding three-dimensional binding motifs and optimizing small molecules for protein interaction.

Hydrogen bond interaction geometries in protein-ligand complexes: From large-scale statistics to single cases

253rd National Meeting (2017)

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.