Cartaj

Cartaj predicts the topological family of RNA three-way junctions from RNA sequence and secondary structure using Watson-Crick pairings to infer junction topology for structural analysis.

Key Features:

• Automated Topology Prediction: Identifies the correct topological configuration among nine possible RNA three-way junction configurations using secondary structure data alone.
• Algorithmic Foundation: Model parameters were determined from a dataset of 33 RNA three-way junctions with known three-dimensional conformations.
• Performance Metrics: Achieved a 64% success rate when tested on an independent set of 53 junctions for correct topological configuration prediction.
• Enhanced Accuracy with Homology Information: Prediction accuracy is significantly improved when homologous sequences are incorporated into the analysis.

Scientific Applications:

• RNA structural analysis: Facilitates classification of three-way junction topologies to support structural studies of RNA molecules.
• Functional interpretation: Provides topological insights relevant to biological processes such as splicing, ribosome assembly, and viral replication.

Methodology:

Input consists of RNA sequence and secondary structure specifying Watson-Crick pairings; a trained algorithm evaluates candidate topological configurations and selects the most probable one based on learned parameters; homologous sequence information can be integrated to refine predictions.