Chemosensory Science
The CPM group aims to provide answers to the following three key questions:
- How do different flavor-active molecules encode for different chemosensory qualities?
- How do chemoreceptors decode this molecular information embedded in our food?
- Can this understanding be leveraged to build accurate predictive models of flavor perception?
Chemoinformatics & AI Prediction
Bitter Peptide Prediction: Our DFG-funded project PI1672/3 focuses on the prediction of peptide bitter taste. We developed BitterPep-GCN, a graph neural network (GNN) predictor for peptide bitterness. The model’s real-world applicability was experimentally validated through human sensory studies in the research group of Prof. Dr. Corinna Dawid and functional assays in the research group of Dr. Maik Behrens.
Protein Structure-Dynamics Modeling
Conformational landscape of odorant receptors: As a spin-off from the GPCRmd consortium, our group pioneered the first complete structural dataset of refined odorant receptor (OR) models using cutting-edge high-throughput molecular dynamics simulations (ORmd project b178bb).
Virtual Screening
Ligand Discovery for Chemosensory Receptors: As part of the Leibniz Women Professors grant (P116/2020), the CPM group has developed a computational toolbox for generating predictive 3D structure models of bitter taste and olfactory receptors. These models have been successfully applied in virtual screening campaigns, and through collaboration, the predictions were experimentally validated, leading to the discovery of novel chemoreceptor modulators.