Monte Carlo basin paving: an improved global optimization method.

We propose a global optimization procedure, basin paving, which is based on the combination of the optimization strategies behind basin hopping and energy landscape paving. As an example, we describe its application in the protein structure prediction by examining two well-studied peptides, where we have found lower potential energy minima than previously located. We also compare the statistics of the searching trajectories produced by basin paving, basin hopping, and energy landscape paving.