Authors: Th. Herges, A. Schug and W. Wenzel
Affilation: Forschungszentrum Karlsruhe, Germany
Pages: 90 - 93
Keywords: protein folding, forcefield development, protein structure prediction
Protein structure prediction remains one of the main outstanding challenges of theoretical biophysical chemistry. Its goal is to predict the fully three-dimensional tertiary structure of the protein on the basis of its amino acid sequence alone. Here we report progress on the development of specific biomolecular forcefield for protein structure prediction, which is parameterised by: -van-der-Waals parameters extracted from the PDB database, -group-specific dielectic constants-environment dependent partial charges -a novel model for backbone-backbone hydrogen bonding and -an implicit area based solvent model. Under the assumption that the protein is in thermodynamic equilibrium with its environment, its native state corresponds to the global minimum of the free-energy surface specified by the forcefield. Using a widely studied example, the 36 amino acid headpiece of the villin protein (1VII) (see Fig 1(a)), we found that our original forcefield predicted non-native structures with free energies below that of the known NMR structure of this autonomously folding peptide. Using a decoy approach we optimised the forcefield by allowing the free energies per surface area to vary by 15% in order to stabilize the native state. In the optimised forcefield a near-native configuration (see Fig 1(b)) became the lowest energy configuration among over 10000 specifically optimised competing decoys. Using this decoy set we were able to completely characterize the tree of low energy configurations of the peptide. We have then applied the same optimisation strategy to two other peptides, the HIV accessory protein (1F4I) (see Fig 2(a)) and XXX (1BH) (see Fig 2(b)). Without further parameter adjustment, near native structures were predicted for both proteins in sets of 4600 and 6000 decoys respectively, which demonstrates a certain degree of transferability of the forcefield. Acknowledgments: This work was funded by the German National Science Foundation (DFG We 1883/11-1), the department of state for science and technology and the Bode Foundation.