UMass Morph Server: Macromolecular Dynamics Analyses Using Elastic Network Models

Abstract

The geometry-based mechanical models called elastic network models (ENMs) in various resolutions have been developed for the study of macromolecular motions. In a coarse-grained ENM, a biological system is represented as a network of springs connecting representative points. They range from single atoms to functional domains depending on the level of details in modeling. In this paper presented are the various kinds of coarse-graining methods such as symmetry-constrained, rigid-cluster, and hybrid ENMs. They enable us to overcome the computational burden and memory limitation in the conventional molecular dynamics (MD) simulations and full-atom normal mode analysis (NMA) without loss of generality. For the broad impact of this work on the structural biology area we also develop the UMass Morph Server (UMMS). Based on the requests from online users, UMMS does not only serve a harmonic NMA that describes thermal behaviors (i.e., fluctuations) of a macromolecule around its equilibrium state, but also generates anharmonic transition pathways between two end conformations by using the elastic network interpolation (ENI) also developed by the author. In addition, UMMS can provide two unique features as follows: (i) interpretation of massive MD data by finding essential pathways (ii) the conformation prediction incorporated with time-resolved information such as FRET data. Many example movies and numeric data can be downloadable at http://biomechanics.ecs.umass.edu/umms.html.