A global optimization strategy for predicting alpha-helical protein tertiary structure

We present a global optimization strategy that incorporates predicted restr aints in both a local optimization context and as directives for global opt imization approaches, to predict protein tertiary structure for cc-helical proteins. Specifically, neural networks are used to predict the secondary s tructure of a protein, restraints are defined as manifestations of the netw ork with a predicted secondary structure and the secondary structure is for med using local minimizations on a protein energy surface, in the presence of the restraints. Those residues predicted to be coil, by the network, def ine a conformational sub-space that is subject to optimization using a glob al approach known as stochastic perturbation that has been found to be effe ctive for Lennard-Jones clusters and homo-polypeptides. Our energy surface is an all-atom 'gas phase' molecular mechanics force field, that is combine d with a new solvation energy function that penalizes hydrophobic group exp osure. This energy function gives the crystal structure of four different c c-helical proteins as the lowest energy structure relative to other conform ations, with correct secondary structure but incorrect tertiary structure. We demonstrate this global optimization strategy by determining the tertiar y structure of the A-chain of the cc-helical protein, uteroglobin and of a four-helix bundle, DNA binding protein. (C) 2000 Elsevier Science Ltd. All rights reserved.