RNANetMotif

RNANetMotif identifies sequence-structure motifs in RNA-binding protein (RBP) binding sites by mining graph representations of predicted RNA secondary structures to reveal enriched network subgraphs relevant to RBP–RNA recognition.

Key Features:

• Graph-based representation: RNA sequences are represented as graphs with nucleotides as nodes and backbone covalent bonds plus base-pairing hydrogen bonds as edges to capture sequence and secondary-structure connectivity.
• Enriched subgraph identification (Extended K-mer Subgraphs, EKS): The method detects enriched subgraphs, termed Extended K-mer Subgraphs (EKS), that correspond to recurrent sequence-structure motifs in RBP-bound RNAs.
• GraphK algorithm: A specialized graph search algorithm, GraphK, is used to efficiently mine the RNA graphs for enriched sequence-structure subgraphs.

Scientific Applications:

• RBP–RNA interaction modeling: Incorporating sequence and structural information to improve models that predict RBP binding sites and specificity.
• Binding specificity elucidation: Revealing sequence-structure elements that explain how RBPs recognize specific RNA targets.
• Functional RNA biology: Informing studies of cellular processes influenced by RBP binding, including splicing, translation, RNA transport, and degradation, and enabling adaptation to other biomolecular sequence-structure interaction problems.

Methodology:

Input consists of predicted secondary structures of thousands of RBP-bound RNA sequences (derived from RNAcompete and eCLIP); sequences are converted to graphs (nucleotides nodes; backbone covalent and base-pairing hydrogen-bond edges); GraphK is applied to identify enriched Extended K-mer Subgraphs (EKS); validation uses discrete molecular dynamics folding simulations and RNA–protein docking for LIN28 assessing solvent accessibility and energetics, and selected RNAs are modeled in 3D for spatial evaluation.