A Silicon Graphics Iris computer system will be acquired for research in computational biology. With this computer system, Professor Levitt will bring computer simulation of protein molecular dynamics to the point where it is sufficiently reliable and realistic to allow prediction of energetic, structural and dynamic properties of these complicated systems. This will be accomplished by asking specific questions and answering them by simulation and detailed analysis of specific applications as follows: 1. Are simulations of proteins in solution sufficiently stable to allow the extended simulation needed to sample rare events and accumulate accurate averages? Extending simulations of bovine pancreatic trypsin inhibitor (BPTI) in solution beyond 100 picoseconds (ps) will confirm their methodological advances (use of all-atom energy functions, a flexible water molecule and a smoothly truncated nonbonded potential together with strict insistence on energy conservation). 2. How does the solution structure of a protein differ from the crystal structure? Analysis of the extended BPTI trajectory by calculating the spectral densities of each inter-proton vector will make it possible to derive a more accurate solution structure from nuclear magnetic resonance experiments. Comparisons to inelastic neutron scattering data will also be made. 3. Can simple energy functions and classical dynamics reproduce the details of protein x-ray structures? They will simulate the dynamics of two well-refined, high-resolution protein structures, BPTI and crambin in their native crystal lattices. Systematic differences between the calculated time-averaged structure and the x-ray structure will be used to modify the energy parameters. 4. How does water penetrate into and disrupt native protein structure in the first stages of unfolding? Which interactions are weakened first, the polar hydrogen bonds or the non-polar van der Waals' interactions? How rapidly do such "unfolding" fluctuations occur under normal and denaturing conditions? They will investigate the stability of native BPTI in solution as temperature, density (equivalent to pressure) and patterns of disulphide bonding are changed. These methods will also be used on simpler systems such as alpha-helices and beta-hairpins in solution. 5. By what path and at what rate are water molecules expelled from protein binding sites? They will apply these methods to larger proteins, ribonuclease and the trypsin/BPTI complex, which are also being studied experimentally.
