Dual graph partitioning highlights a small group of pseudoknot-containing RNA submotifs

Swati Jain, Cigdem S. Bayrak, Louis Petingi, Tamar Schlick

Research output: Contribution to journalArticlepeer-review

Abstract

RNA molecules are composed of modular architectural units that define their unique structural and functional properties. Characterization of these building blocks can help interpret RNA structure/function relationships. We present an RNA secondary structure motif and submotif library using dual graph representation and partitioning. Dual graphs represent RNA helices as vertices and loops as edges. Unlike tree graphs, dual graphs can represent RNA pseudoknots (intertwined base pairs). For a representative set of RNA structures, we construct dual graphs from their secondary structures, and apply our partitioning algorithm to identify non-separable subgraphs (or blocks) without breaking pseudoknots. We report 56 subgraph blocks up to nine vertices; among them, 22 are frequently occurring, 15 of which contain pseudoknots. We then catalog atomic fragments corresponding to the subgraph blocks to define a library of building blocks that can be used for RNA design, which we call RAG-3Dual, as we have done for tree graphs. As an application, we analyze the distribution of these subgraph blocks within ribosomal RNAs of various prokaryotic and eukaryotic species to identify common subgraphs and possible ancestry relationships. Other applications of dual graph partitioning and motif library can be envisioned for RNA structure analysis and design.

Original languageEnglish (US)
Article number371
JournalGenes
Volume9
Issue number8
DOIs
StatePublished - Aug 2018

Keywords

  • Dual graphs
  • Graph partitioning
  • Pseudoknots
  • RNA graphs
  • RNA substructures and submotifs
  • Ribosomal RNAs

ASJC Scopus subject areas

  • Genetics
  • Genetics(clinical)

Fingerprint Dive into the research topics of 'Dual graph partitioning highlights a small group of pseudoknot-containing RNA submotifs'. Together they form a unique fingerprint.

Cite this