ATLAS: Graph-based 3D RNA Motif Library Incorporating non-Watson-Crick Interactions

Area of Science:

• Structural Biology
• Bioinformatics
• Computational Biology

Background:

• RNA motifs are recurring structural patterns essential for RNA function.
• Existing RNA 3D structure libraries lack comprehensive representations of non-Watson-Crick interactions and pseudoknots.

Purpose of the Study:

• To develop the Advanced Template Library for Assembly and Structure (ATLAS) for RNA motif analysis.
• To provide a resource for RNA 3D structure prediction and design, incorporating non-Watson-Crick interactions.
• To establish a measure of RNA structural similarity based on shared motifs.

Main Methods:

• Developed graph representations for RNA structures, including non-Watson-Crick interactions and pseudoknots.
• Integrated atomic 3D structures from the Protein Data Bank (PDB) into ATLAS.
• Implemented an isomorphism graph searching algorithm for motif identification.
• Created a web service for searching and downloading RNA motifs and user-defined structures.
• Proposed a physics-inspired model for RNA evolution to analyze structural similarity.

Main Results:

• ATLAS provides nucleotide-level graph representations of RNA motifs and corresponding PDB IDs.
• The library includes 3D atomic structures and supports non-Watson-Crick interactions and pseudoknots.
• A novel measure of RNA similarity based on shared motifs was developed.
• The proposed evolutionary model's steady-state solution aligns with observed RNA structural similarity distributions.

Conclusions:

• ATLAS is a comprehensive resource for RNA motif analysis, prediction, and design.
• The inclusion of non-Watson-Crick interactions and pseudoknots enhances RNA structure representation.
• The developed similarity measure and evolutionary model offer new insights into RNA structural diversity and evolution.