Interpreting NMR dynamic parameters via the separation of reorientational motion in MD simulation

Abstract

Reorientational dynamics-motion defined by changes in the direction of a vector or tensor-determine relaxation behavior in nuclear magnetic resonance (NMR). However, if multiple processes exist that result in reorientation, then analyzing the net effects becomes a complex task, so that one ideally would separate those motions to simplify analysis. The model-free and two-step approaches have established the separability of the total correlation function of reorientation motion into contributions from statistically independent motions. Separability has been used to justify the analysis of experimental relaxation rate constants by fitting data to a total correlation function resulting from the product of two or three individual correlation functions, each representing an independent motion. The resulting parameters are used to describe motion in the molecule, but if multiple internal motions are present, interpreting those parameters is not trivial. We suggest an alternative approach: quantitative and timescale-specific comparison of experiment and simulation, as previously established using the detector analysis. This is followed by separation of simulated correlation functions into independent motions, and timescale-specific parameterization of the results, such that one may determine how each motion contributes to experimental parameters. We establish protocols for the separation of the correlation function into components using coordinates from molecular dynamics simulation. Separation is achieved by defining a series of frames, where the frames iteratively split the total motion into contributions from motion within each frame and of each frame. Then timescale specific parameters (e.g. detector responses) describing the total motion may be interpreted in terms of the timescale-specific parameterization of the individual motions.