Michael R. Shirts1, Guha Jayachandran2, Christopher D. Snow3, Vijay S. Pande4, and Richard A. Friesner1. (1) Department of Chemistry, Columbia University, Havemeyer Hall MC 3131, New York, NY 10027, (2) Computer Science, Stanford University, Clark Center MC 9025, Stanford, CA 94305, (3) Chemical Engineering, California Institute of Technology, M/C 210-41, 91125, Pasadena, CA, (4) Department of Chemistry, Stanford University, MC 5080, Stanford, CA 94305-5080
Recent developments in computational processing power [1] and in theoretical methods [2] have made it possible to compute absolute ligand binding free energies to a precision of under 1 kcal/mol. This level of precision makes it possible to separate errors in sampling from inadequacies of molecular mechanics force fields, and begins to bring the design of drugs with high affinities from an art to a process of nanoscale engineering.
We compare experimental free energies of binding of FKBP-12 with a variety of ligands with extensive explicit solvent absolute free energies via molecular dynamics, with converged calculations being within 1 kcal/mol of experiment. We also present advances in continuum solvent Monte Carlo methods for calculating ligand binding affinity on this and other systems.
[1] M. Shirts and V. S. Pande, Science, 290:1903 (2000)
[2] M. R. Shirts and V. S. Pande, J. Chem. Phys. 122:134508 (2005)
Web Page:
www.columbia.edu/~mrs2144/MRSCV.html