TOWARDS AN EXHAUSTIVE SAMPLING OF THE CONFIGURATIONAL SPACES OF THE 2FORMS OF THE PEPTIDE-HORMONE GUANYLIN

The recently introduced Essential Dynamics sampling method is extended such that an exhaustive sampling of the available (backbone) configur ational space can be achieved. From an initial Molecular Dynamics simu lation an approximated definition of the essential subspace is obtaine d. This subspace is used to direct subsequent simulations by means of constraint forces. The method is applied to the peptide hormone guanyl in, solvated in water, of which the structure was determined recently. The peptide exists in two forms and for both forms, an extensive samp ling was produced. The sampling algorithm fills the available space (o f the essential coordinates used in the procedure) at a rate that is a pproximately six to seven times larger than that for traditional Molec ular Dynamics. The procedure does not cause any significant perturbati on, which is indicated by the fact that free Molecular Dynamics simula tions started at several places in the space defined by the Essential Dynamics sample that complete space. Moreover, analyses of the average free Molecular Dynamics step have shown that nowhere except close to the edge of the available space, there are regions where the system sh ows a drift in a particular direction. This result also shows that in principle, the essential subspace is a constant free energy surface, w ith well-defined and steep borders, in which the system moves diffusiv ely. In addition, a comparison between two independent essential dynam ics sampling runs, of one form of the peptide, shows that the obtained essential subspaces are virtually identical.