Statistical Physics of Protein Folding and Aggregation

Abstract: The mechanisms of protein folding and aggregation are investigated by computer simulations of all-atom and reduced models with sequence-based potentials. A quasi local Monte Carlo update is developed in order to efficiently sample proteins in the folded phase. A small helical protein, the B-domain of staphylococcal protein A, is studied using a reduced model. In the thermodynamically favoured topology, energy minimisation leads to a conformation whose root mean square deviation form the experimental structure is 1.8Å. We also study the thermodynamics and kinetics of small fast folding proteins without a clear free-energy barrier between the folded and unfolded states. Analytical calculations using a square well-potential enable us to predict the relaxation time within a factor of two. Finally using an all atom model, we study the aggregation properties of a 7-amino acid fragment of Alzheimer's amyloid beta peptide. We find that the system of three and six such fragments form aggregated structures with a high content of antiparallel beta-sheet structure, which is in line with experimental data.