Proteins have been shown to exhibit strange/anomalous dynamics displaying non-Debye density of vibrational states, anomalous spread of vibrational energy, large conformational changes, nonexponential decay of correlations, and nonexponential unfolding times. The anomalous behavior may, in principle, stem from various factors affecting the energy landscape under which a protein vibrates. Investigating the origins of such unconventional dynamics, we focus on the structure-dynamics interplay and introduce a stochastic approach to the vibrational dynamics of proteins. We use diffusion, a method sensitive to the structural features of the protein fold and them alone, in order to probe protein structure. Conducting a large-scale study of diffusion on over 500 Protein Data Bank structures we find it to be anomalous, an indication of a fractal-like structure. Taking advantage of known and newly derived relations between vibrational dynamics and diffusion, we demonstrate the equivalence of our findings to the existence of structurally originated anomalies in the vibrational dynamics of proteins. We conclude that these anomalies are a direct result of the fractal-like structure of proteins. The duality between diffusion and vibrational dynamics allows us to make, on a single-molecule level, experimentally testable predictions. The time dependent vibrational mean square displacement of an amino acid is predicted to be subdiffusive. The thermal variance in the instantaneous distance between amino acids is shown to grow as a power law of the equilibrium distance. Mean first passage time analysis is offered as a practical tool that may aid in the identification of amino acid pairs involved in large conformational changes.protein dynamics | Gaussian network model | anomalous diffusion | intramolecular distance distributions | mean first passage time
IntroductionProteins are commonly envisioned vibrating under the influence of a complicated energy landscape (1) held responsible for their intricate dynamics (1-9). Despite its overall complexity the energy landscape is approximately harmonic near its minima, a fact that does not seem to coincide with the rich dynamics mentioned above. Attacking the validity of the harmonic approximation is one way out of this puzzle. One might have suggested that the harmonic approximation is unable to capture the anomalous dynamics displayed by proteins since the latter is dominated by large excursions from the native state structure. The failure of the harmonic approximation then follows from the fact that it does not describe the energy of conformations which considerably deviate from the native state structure. In this paper we argue differently showing that the native state structure of proteins is fractal-like and hence anomalous in its basis. As a result one needs not go beyond the harmonic approximation in order to observe anomalous dynamics since anomalous structure leads to it even in the presence of the simplest interactions. Interestingly, when the structure is fractal-like, large con...