Statistical Analysis of Structural Transitions in Small Systems

We discuss general thermodynamic properties of molecular structure formation processes like protein folding by means of simplified, coarse-grained models. The conformational transitions accompanying these processes exhibit similarities to thermodynamic phase transitions, but also significant differences as the systems that we investigate here are very small. The usefulness of a microcanonical statistical analysis of these transitions in comparison with a canonical interpretation is emphasized. The results are obtained by employing sophisticated generalized-ensemble Markov-chain Monte Carlo methodologies.

💡 Research Summary

The paper investigates structural transitions in very small molecular systems, with a focus on protein folding, using simplified coarse‑grained models and advanced statistical‑mechanical analysis. The authors begin by reviewing the conventional thermodynamic description of polymer and protein phase behavior, emphasizing the well‑known scaling laws for long chains (ν≈3/5 in good solvent, ν=1/2 at the Θ‑point, ν≈1/3 in the collapsed globular state). They point out that these asymptotic results rely on the thermodynamic limit (N→∞) where surface effects become negligible. In contrast, real polymers of interest—especially those relevant to nanotechnology and modern biophysics—have finite lengths (N on the order of tens to a few hundred monomers). For such finite systems, the surface‑to‑volume ratio remains large, and the notion of a sharp phase transition breaks down. Instead, one observes broad temperature intervals over which the system changes its dominant conformational ensemble; the authors refer to these as “pseudophase transitions”.

To explore these finite‑size effects, the study employs two representative coarse‑grained models. The first is a lattice peptide consisting of 42 residues that incorporates hydrophobic–polar (HP) interactions and a simplified hydrogen‑bonding term. Using multicanonical Monte Carlo simulations, the authors obtain an accurate estimate of the density of states g(E). From g(E) they compute the microcanonical entropy S(E)=ln G(E) (with G(E) the integrated density of states) and the associated caloric temperature T(E)=