Graph-based sampling for approximating global helical topologies of RNA

Abstract

A current challenge in RNA structure prediction is the description of global helical arrangements compatible with a given secondary structure. Here we address this problem by developing a hierarchical graph sampling/data mining approach to reduce conformational space and accelerate global sampling of candidate topologies. Starting from a 2D structure, we construct an initial graph from size measures deduced from solved RNAs and junction topologies predicted by our data-mining algorithm RNAJAG trained on known RNAs. We sample these graphs in 3D space guided by knowledge-based statistical potentials derived from bending and torsion measures of internal loops as well as radii of gyration for known RNAs. Graph sampling results for 30 representative RNAs are analyzed and compared with reference graphs from both solved structures and predicted structures by available programs. This comparison indicates promise for our graph-based sampling approach for characterizing global helical arrangements in large RNAs: graph rmsds range from 2.52 to 28.24 Å for RNAs of size 25-158 nucleotides, and more than half of our graph predictions improve upon other programs. The efficiency in graph sampling, however, implies an additional step of translating candidate graphs into atomic models. Such models can be built with the same idea of graph partitioning and build-up procedures we used for RNA design.RNA 3D graph | Monte Carlo simulated annealing | RNA 3D prediction T he heightened interest in RNA biology with demonstrated successful applications to medicine and technology has presented new challenges to computational scientists in RNA structure prediction. Though general automated prediction of RNA tertiary (3D) structure from the primary sequence remains elusive, many effective approaches exist for analyzing and describing 3D RNA structures as well as predicting reasonably 3D aspects of small RNAs, ranging from coarse-grained modeling (1) to various structure assembly (2), energy minimization (3), molecular dynamics (4), and other conformational sampling approaches (5, 6).Interest in RNA structure prediction and its modular architecture has also led to many analyses of RNA local structure (7-12). In particular, several studies have focused on the helical arrangements formed by internal loops, important points of flexibility that can affect the overall 3D shape of RNAs. Indeed, the bending and torsion of helical arms connected by internal loops define unique helical conformations, as analyzed by AlHashimi and coworkers (7), Tang and Draper (8), Hagerman and coworkers (9), and Olson and coworkers (10). Recently, Pyle and coworkers (11) reported a pseudotorsional angle database from local RNA backbone geometry, and Sim and Levitt (12) cataloged preferred helical arrangements among nucleotide fragment assemblies given a secondary (2D) conformation. However, extensive topological and geometrical analyses over a large diverse set of RNAs do not exist.To such endeavors, mathematical and computational tools have been applied, including gra...

Predicting RNA 3D structure using a coarse-grain helix-centered model

“…Whereas previous attempts at reducing the degrees of freedom in an RNA molecule have ranged from using three points to represent a nucleotide (Ding et al 2008), to using one point to represent a nucleotide (Jonikas et al 2009), we represent the helix using one line segment and two vectors and consider elements linking helices as the degrees of freedom. It should be noted that a recent approach (Kim et al 2014) has presented a very similar model using a helix-as-astick representation of RNA 3D structure and combining it with predictions of local junction topology to provide accurate predictions of RNA structures. While our approaches overlap in the abstraction of the structure, our method for sampling local structure as well as our energy function formulations differ significantly.…”

“…It is derived from RNA secondary structures and defines the structural relations of individual helices. Similar graph representations and their use in structure prediction have been mentioned by Zhao et al (2012), Lamiable et al (2013), and Kim et al (2014) but we aim to formalize their definition and illustrate its use as a guide for building a coarse-grain 3D structure. II.…”

Progress and Current Challenges in Modeling Large RNAs

“…For example, NAST [51] and YUP [52] represent each nucleotide as a single pseudo atom, Vfold [53], SimRNA [54] and iFoldRNA [55] define each nucleotide using three pseudo atoms, CG [56] represents each nucleotide with five pseudo atoms, and HiRE-RNA [57] uses six pseudo atoms to represent purines and seven for pyrimidines. In addition, there are methods that represent RNA as graphs, which further reduces the sampling space by representing each helix as a stick or a cylinder [58, 59]. The accuracy of coarse-grained methods depends on the choice of representation and the scoring function.…”

F-RAG: Generating Atomic Coordinates from RNA Graphs by Fragment Assembly

“…Recently, we have developed a hierarchical graph sampling methodology, called RAGTOP (RNA-As-Graphs Topology Prediction), to predict RNA 3D graph topologies corresponding to a given RNA 2D structure [50]. Our Junction-Explorer data mining program [44, 45] is first used to determine the junction orientation (co-axial stacking and family) of the candidate sequence and 2D structure, as classified in our junction analysis work.…”