Prediction of Protein-Ligand and Protein-Protein Interactions
Faculty: Vicente Reyes
Protein-ligand interactions (PLIs) and protein-protein interactions (PPIs) are all-important recurring themes in biology. We have developed a method to model ligand binding sites and protein-protein interactions sites using a reduced protein representation. In PLIs the binding site is modeled as a tetrahedral motif composed of amino acid centroids, and this model is extended as a pair of interacting tetrahedra in PPIs. We have also developed a search algorithm to determine the presence of these tetrahedral motifs in protein 3D structures. These set of procedures thus allow structure-based functional annotation of novel proteins whose structures are known (experimental or predicted). We are extending this computational modeling approach to other types of biomolecules such as nucleic acids (DNA and RNA), carbohydrates (sugars) and lipids (fats) in an attempt to systematize structure-based function prediction as well as structural and dynamical modeling of all four main types of biomolecules.
Link: Protein DCRR Web Server