On coarse-grained Normal Mode Analysis and refined Gaussian Network Model for protein structure fluctuations

摘要: The functions of biological structures are related to the dynamics structures, especially various kinds large-amplitude molecular motions. With some assumptions, those motions can be investigated by Normal Mode Analysis (NMA) and Gaussian Network Model (GNM). However, despite their contributions many applications, relationship between NMA GNM requires a further discussion. In this work, we review address common applications in structural biology. We evaluate GNM, based on how well it predicts fluctuations, compared experimental data for large set protein structures. Then, propose several ways coarse-graining residue-level fluctuations choosing different approaches represent amino acids forces them. Using backbone atoms such as Cα, C, N , Cβ, single-atom representations considered. Combinations these also tested representative point residue. force constants extracted from Hessian matrix potential energy used corresponding residues. residue mean-square-fluctuations correlations with B-factors calculated proteins. results all-atom normal mode analysis choice atoms. coarse-grained methods perform more efficiently than analysis, agree better B-factors. B-factor comparable or estimated conventional GNM. surveyed pairs residues extents separation sequence. statistical averages build finer-grained here called nonhomogeneous which is able predict mean square significantly test cases.