The principal objectives of this laboratory are to develop and apply computational and computer graphic techniques in the study of biomolecular interactions. The growing data base of three-dimensional protein and nucleic acid structures highlights the need for new methods of analyzing and predicting molecular interaction. Our focus is on protein-protein recognition and binding in oligomer formation, antibody-antigen complexes, and on protein-substrate interactions in drug design. A new laboratory initiative is in developing methods to model "From Atoms to Cells."
We have developed a number of techniques for visualizing and analyzing protein structure and properties. We have also recently developed new representations of protein surfaces and properties using an expansion of spherical harmonic functions. This new characterization enables a flexible and continuous description of molecules over multiple length scales. We have used this approach to develop SurfDock-a program to predict protein-protein association.
We are continuing development of techniques to predict protein interactions. Our AutoDock code uses efficient global and local search procedures to dock flexible substrates into protein active sites. We are expanding these techniques to the design of new inhibitors, focusing on anti-HIV therapeutics, and the blood coagulation cascade.