The cell is highly dynamic in character, with all observable behaviors and attributes governed by kinetic interactions of its macromolecular elements. Among these interactions, formation of both transient and stable protein-protein complexes rank among both the most important and the least understood. Central to the understanding of protein-protein interactions has been the characterization and analysis of protein-protein interfaces. While numerous computational studies have added to our knowledge of protein-protein interfaces, little of this information has resulted in better prediction of interface sites, improvement in protein-protein docking attempts, or good classifications of types of interfaces. We believe that the primary reason for this shortcoming is that the interfaces used in studies thus far have been generated too arbitrarily, preventing the application of rigorous measures. As a result, the only descriptors that can be used are statistical or geometrically too coarse to capture and discriminate structure embedded in the interface.

We have developed an interface construct based upon concepts surrounding the Alpha Shape theory developed by Edelsbrunner and collaborators over the past decade, that alleviates problems encountered when using traditional ways of generating protein-protein interfaces. The construct is a surface, and is rigorously defined by geometric and topological principles. Software implementation is currently underway, as is investigation of characterizing the interface by using classical biochemical descriptors (force-field style energies) and development of new geometric and topological descriptors. After the correlation with existing, published data on protein interfaces, we hope to experimentally validate our definition by using the interface to direct studies of a nontrivial and transient protein complex. Currently we are considering various systems with a focus on enzymes and their natural protein substrates involved in cell cycle progression, such as the phosphatase Cdc25B and its natural substrate, the cyclin dependent kinase-cyclin complex, Cdk2/Cyclin A. Possible experimental techniques to employ include fluorescence as the tool of choice for measuring binding kinetics and ITC to measure binding energies.