Activity profiling and prediction of off-pharmacologies
Please notice these positions are no longer vacant.

Molecular basis of the interactions between the hERG channel and its ligands.
Interaction with the hERG potassium channel comprises a severe safety risk for compounds under development. Thus, computational models which allow the prediction of hERG binding of structurally diverse compound libraries and which provide evidence for the molecular basis of drug/channel interaction are of great interest. Within this project we aim at the development and validation of ligand- and structure-based in silico models for prediction of drug/hERG channel interaction. Ligand-based design approaches such as 2D- and 3D-QSAR, use of GRIND/ALMOND and pharmacophore modelling will lead to QSAR models which should allow the identification of molecular features relevant for hERG binding. We will utilize both our in house derived library of compounds extracted from literature as well as data on a set of propafenone analogues obtained in the lab of S. Hering. Structure-based design methods will mainly rely on docking of compound libraries into homology models of the open and closed states of the hERG channel utilizing MOE, GOLD, FlexE/FlexX and Prime/Glide software packages. Due to the promiscuity of the hERG channel with respect to ligand recognition and the lack of available X-ray structures, the models will be subject to rigorous validation based on experimental data and biophysical studies (E. Timin, S. Hering).

Selectivity profiling of GABAA and TRPV1 receptor ligands.
Piperine has recently been identified as a modulator of GABAA receptors and was also shown to interact with TRPV1 receptors. Piperine will serve as an interesting example to elucidate the target (GABAA vs. TRPV1) specific properties of the basic scaffold and to perform selectivity profiling studies. Within the project, we will develop ligand-based in silico models for designing in/designing out interaction of piperine analogues with GABAA and TRPV1 receptors. Methods applied will include conventional 2D-QSAR techniques such as Hansch analysis and Free-Wilson analysis as well as 3D-methods such as CoMFA/CoMSIA and pharmacophore modelling. These models will guide the chemical synthesis performed in the group of M. Mihovilovic and also provide insights into the molecular features important for drug/receptor interaction. For identification of new scaffolds we will use similarity search as well as application of our recently published self-organising map approach. Descriptors used include typical 2D- and 3D-descriptors, autocorrelation vectors, GRIND/ALMOND descriptors, pharmacophore fingerprints as well as our SIBAR descriptors. Libraries screened will include ChemBridge, ChemDiv, ZINC, Enamine, and others. Virtual screening runs are expected to provide a hit list of potential active compounds with chemical structures different from the piperine scaffold, which are subject to pharmacological testing ion the group of S. Hering.