Abstract Our research focuses on molecular recognition, with the goal of providing methods and software for solving biomedical problems. The primary areas of interest are protein-protein interactions and the ligand binding properties of proteins. We believe that predictive methods will be substantially improved during the next five years due to the increasing amount of information on sequences, structures, and interactions of molecules in the cell, and the unprecedented availability of computing power. To take advantage of these opportunities we will integrate the use of structural templates, co-evolutionary information, and machine learning into classical biophysical methods. Our rigid body protein docking server ClusPro, which has over 15,000 users, will be combined with our new template based server ClusPro TBM. We also add elements of flexible docking, either by remodeling the regions that cause steric conflicts, or by using a neural net for calculating post-minimization energy values without performing the actual minimization. Several tools will be combined for the structural analysis of protein interaction networks, including a novel method of constructing multi-protein complexes based on pre-calculated tables of interaction energies between pairs of proteins. Examples of applications include the design of PROteolysis TArgeting Chimeras (PROTACs) for modulating a target protein by degradation, the prediction of antibody epitopes, and searching for epitope-specific antibodies. To study the ligand binding properties of proteins we focus on binding hot spots, regions of proteins that are major contributors to the binding free energy. Our FTMap server globally samples the surface of target proteins using fragment sized molecular probes and provides reliable hot spot and pharmacophore information. We will improve the scoring function using neural nets, and expand the set of probes to obtain generalized pharmacophores that identify regions in the protein binding site with preferences for specific functional groups and a number of bound fragments. Since this information can be used to find larger ligands, the goal is to convert FTMap into a fragment based ligand discovery platform. We will also improve our template-based server LigTBM, which docks small molecules to proteins, and will integrate template-based modeling with FTMap. In a collaborative application we will analyze metabolite-protein interaction data obtained by precision mass spectrometry in E. coli and human protein pull-down experiments. FTMap will be used to test whether a target protein has a suitable binding hot spot, and LigTBM will place the metabolite. We are particularly interested in finding metabolites that bind at novel allosteric regulatory sites. A related application will be to study ensembles of structures obtained by dynamic simulations to find potential correlations between FTMap derived binding properties at different regions of proteins, thus exploring potential allosteric communication.

Narrative Our research focuses on molecular interactions, and we develop methods for protein-protein docking and the characterization of ligand binding properties of proteins. Predictive tools for these applications will be substantially improved by integrating classical biophysical approaches with sequence, structure, and interaction information on biological macromolecules, and by utilizing the unprecedented availability of computer power. The methods will be implemented as web-based servers for use by the biomedical research community.