The density of states contains all informations on energetic quantities of a statistical system, such as the mean energy, free energy, entropy, and specific heat. As a specific application, we consider in this work a simple lattice model for heteropolymers that is widely used for studying statistical properties of proteins. For short chains, we have derived exact results from conformational enumeration, while for longer ones we developed a multicanonical Monte Carlo variant of the nPERM-based chain growth method in order to directly simulate the density of states. For simplification, only two types of monomers with respective hydrophobic (H) and polar (P) residues are regarded and only the next-neighbour interaction between hydrophobic monomers, being nonadjacent along the chain, is taken into account. This is known as the HP model for the folding of lattice proteins.

PACS numbers: 05.10.--a, 87.15.Aa, 87.15.Cc