![]() |
MIT Electrical Engineering and Computer Science
EECS Event |
Thursday, March 1, 2001
4:15 PM (refreshments 4:00)
Room NE43-518
EECS Special Seminar
Abstract
The development of algorithms to optimize packing interactions in molecules has led to dramatic advances in protein design. Progress in the field of structure-based rational drug design has been much slower due to the need not only for complementary packing interactions but also complementary electrostatic interactions between drug and protein. Algorithms for designing such complementary electrostatic interactions have been lacking. Because binding occurs in water solutions, the large desolvation penalty paid by interfacial polar and charged groups upon binding is counterbalanced by attractive interactions recovered in the bound state. The design problem is to discover drug charge distributions for which this balance leads to the most favorable net contribution to binding. A novel procedure has been developed to solve this problem. Its application to a number of biological systems illustrates the importance of optimized electrostatics in tight-binding protein complexes, the accuracy of the underlying continuum models, and the ability of the approach to attack real-world design and analysis problems.