Identifying Local Regions of Order and Disorder in FG-Nucleoporins and Partially Disordered Proteins Using Molecular Dynamics Simulations

Abstract

Analysis tools optimized to study molecular dynamics (MD) simulations of intrinsically disordered proteins (IDPs) are currently rare. We are developing a set of tools using the libraries and interface of the widely used MD simulation software Gromacs 4 [Hess, et al. 2008]. We are presenting the results of one of these tools which shows utility in differentiating between ordered and disordered regions of proteins based on MD simulation trajectories. This tool assigns a value along a spectrum of order to disorder to a protein based on a scaled average of all inter-structure distances and is based on a previously proposed and intuitive algorithm [Stultz, et al. 2011]. The primary improvements made to this algorithm include additional options of inter-structure distance metrics, a feature to assign values of disorder to individual amino acids, and an output format similar to currently available disorder prediction tools which use amino acid sequence information. When applied to systems of highly disordered protein, this tool is sometimes able to differentiate the regions of secondary structure formed during MD simulations. We are currently studying applications to partially disordered proteins to simplify the identification of local ordered and disordered regions from MD simulation trajectories. This method also provides multiple options of inter-structure distance metrics to optimize the analysis of both mostly ordered and highly flexible proteins. We are validating this usage of the tool with (1) a coarse grain model of several FG-nucleoporin sequences believed to be IDPs with varying levels of disorder spread across different regions and (2) all atom simulations of partially disordered proteins including a fragment of the tumor suppressor protein p53.