rPredictorDB

rPredictorDB predicts RNA secondary structures using experimentally derived templates and visualizes large and complex RNAs to support analysis of RNA structure and function.

Key Features:

• Template-Based Prediction: Employs a template-based approach that leverages experimentally identified structures to predict secondary structures for homologous RNA families, with a repository of 7365 RNAs.
• Integration with Public Resources: Sources sequences from public sequence databases to enable comprehensive structural annotation.
• User-Provided Sequence Prediction: Predicts secondary structures for user-supplied RNA sequences by mapping them to stored templates.
• Advanced Visualization with Traveler: Uses the Traveler software to visualize target RNA secondary structures by applying existing layout templates and transforming tree representations via tree edit distance and layout modification operations.
• Validation and Robustness: Validated against experimentally identified structures and reported to outperform classical de novo prediction algorithms and constrained methods in accuracy and speed.

Scientific Applications:

• Structural studies in molecular biology, genetics, and bioinformatics: Provides predicted and visualized RNA secondary structures for comparative and functional analyses.
• RNA function and identity inference: Facilitates exploration of RNA function and identity through structural annotation and comparison to experimental templates.
• Analysis of ribosomal and large RNAs: Enables analysis and visualization of ribosomal RNAs and other large, complex RNA molecules.

Methodology:

Sequences are sourced from public databases; secondary-structure prediction uses a template-based approach with experimentally identified structures as templates (database of 7365 RNAs); user-provided sequences are predicted by mapping to stored templates; visualization uses Traveler with existing layout templates and transforms tree representations using tree edit distance and layout modification operations; validation was performed against experimentally identified structures and compared to classical prediction algorithms and constrained methods.