Force-constant-decayed anisotropic network model: An improved method for predicting RNA flexibility

Abstract

RNA is an important biological macromolecule, which plays an irreplaceable role in many life activities. RNA functions are largely determined by its tertiary structure and the intrinsic dynamics encoded in the structure. Thus, how to effective extract structure-encoded dynamics is of great significance for understanding RNA functions. Anisotropic network model (ANM) is an efficient method to investigate macromolecular dynamical properties, which has been widely used in protein studies. However, the performance of the conventional ANM in describing RNA flexibility is not as good as that on proteins. In this study, we proposed a new approach, named force-constant-decayed anisotropic network model (fcd-ANM), to improve the performance in investigating the dynamical properties encoded in RNA structures. In fcd-ANM, nucleotide pairs in RNA structure were connected by springs and the force constant of springs was decayed exponentially based on the separation distance to describe the differences in the inter-nucleotide interaction strength. The performance of fcd-ANM in predicting RNA flexibility was evaluated using a non-redundant structure database composed of 51 RNAs. The results indicate that fcd-ANM significantly outperforms the conventional ANM in reproducing the experimental B-factors of nucleotides in RNA structures, and the Pearson correlation coefficient between the predicted and experimental nucleotide B-factors was distinctly improved by 21.05% compared to the conventional ANM. Fcd-ANM can serve as a more effective method for analysis of RNA dynamical properties.