supernmotifs

supernmotifs performs alignment-free comparison of RNA secondary structures to identify and represent structural motifs for scalable analysis of RNA structural patterns.

Key Features:

• Alignment-Free Approach: Employs an alignment-free method to compare RNA secondary structures without sequence alignment.
• Super-n-Motifs Model: Uses a super-n-motifs model with latent analysis to identify enhanced motifs and capture adjacency relations within structures.
• Vector Representation: Computes vector representations of RNA secondary structures as linear combinations of identified motifs.
• Structural Scope: Analyzes both linear and circular RNAs and accommodates complex elements such as pseudoknots and G-quadruplex (G4) motifs.
• Scalability and Efficiency: Performs large-scale comparisons of up to 10,000 secondary structures with reported efficiency gains of up to four orders of magnitude.
• Comparative Representation Advantage: Provides improved representation compared to ordered tree, arc-annotated, and string representations for capturing critical RNA structural features.
• Implementation: Implemented in C++ for computational performance.
• High-Throughput Data Compatibility: Applicable to datasets generated by high-throughput probing techniques.

Scientific Applications:

• Structural Motif Discovery: Identification of recurring and enhanced structural motifs within RNA secondary structures.
• Comparative Structural Analysis: Large-scale comparison and clustering of RNA secondary structures across diverse datasets.
• RNA Function Studies: Investigation of structural patterns relevant to gene regulation and RNA-mediated biological processes.
• Interaction and Therapeutic Research: Analysis supporting studies of RNA-protein interactions and development of RNA-based therapeutics.

Methodology:

Uses an alignment-free computational approach implementing a super-n-motifs model with latent analysis to identify motifs and adjacency relations, computes vector representations as linear combinations of motifs, supports linear and circular RNAs including pseudoknots and G-quadruplex (G4) motifs, and reports large-scale comparison capability (up to 10,000 structures) with up to four orders of magnitude efficiency improvement; implemented in C++.