Molecular Simulation and Systems Biology - Systems Biology: Networks, Models, and Applications - page 82

lower free energy structures. It is important to realize that folding is
a stochastic, rather than a directed, process—a partially folded pro-
tein need not proceed directly to the folded state, but might unfold
partially or completely before attaining a stable structure. In addition,
folded proteins undergo continuous thermal fluctuations that result in
periodic local denaturation and global unfolding events. These events
are difficult to study experimentally, but are clearly observable in
molecular simulations.
More recently, molecular dynamics simulations have been used to
study peptide folding in a predictive fashion. Simulations have been
used to predict the relative rates of folding for different bba mutants
[35] as well as to predict the structures of novel peptide sequences [36].
The continuous advances in computer power suggest that similar
successes are on the horizon for systems as large as a typical protein
(100-300 amino acids).
The Beta Hairpin Peptide from Protein G: A Case Study for the
Interaction Between Experiment, Simulation, and Theory
The 16 C-terminal amino acids of the B1 domain of protein G form a beta
hairpin structure—two strands connected by a turn. (See figure 3.8,
Figure 3.8 Three different views of the 16 amino acid C-terminal beta hairpin
from the B1 domain of protein G. Counterclockwise, from the upper left, the
images show a stick representation, a mixed stick and ribbon representation,
and a ribbon representation with transparent side chains and nearby water
molecules. The striped cylinders show native and nonnative hydrogen bonds
between the two strands.