BkTree

BkTree predicts comprehensive nucleotide interactions to support RNA 3D structure prediction and characterization of non-coding RNA tertiary conformations.

Key Features:

• Comprehensive interaction prediction: Predicts base-base, base-phosphate, base-ribose, and base-stacking interactions, including canonical Watson-Crick and non-canonical pairs.
• Backbone k-tree graph model: Employs a backbone k-tree graph model to constrain nucleotide interaction relationships during prediction.
• Efficient interaction identification: Identifies optimal nucleotide interactions without exhaustive random sampling of conformations.
• Long RNA applicability: Demonstrated capability to predict nucleotide interactions for RNAs longer than 50 nucleotides.

Scientific Applications:

• RNA 3D structure prediction: Supports prediction of tertiary RNA structures by determining detailed nucleotide interaction networks.
• Non-coding RNA analysis: Enables characterization of non-coding RNA tertiary conformations relevant to regulatory functions.
• Improving predictions for longer sequences: Addresses prediction challenges for RNAs exceeding 50 nucleotides where existing methods struggle.

Methodology:

Uses a backbone k-tree graph model to constrain nucleotide interaction relationships and predict base-base, base-phosphate, base-ribose, and base-stacking interactions (canonical and non-canonical), enabling identification of optimal interactions without exhaustive random sampling of conformations.