3D-Jury

3D-Jury predicts protein three-dimensional structures by integrating and scoring structural models generated by multiple protein structure prediction servers using a consensus-based meta-prediction approach.

Key Features:

• Consensus Meta-Prediction Framework: Aggregates structural models from multiple prediction servers to generate consensus protein structure predictions with improved accuracy.
• Meta-Profile Alignment: Utilizes meta-profiles combining predicted secondary structure information and sequence conservation data to enhance detection of structural similarity.
• Confidence Scoring System: Produces confidence scores that correlate with model accuracy and enable estimation of correctly predicted residues.
• Remote Fold Recognition: Detects structural relationships for proteins identifiable by fold recognition but not by PSI-BLAST sequence similarity searches.

Scientific Applications:

• Protein Structural Annotation: Supports structural characterization of proteins lacking experimentally determined structures.
• Remote Homology Detection: Identifies distant evolutionary relationships between protein families through structure-based prediction.
• Structural Biology Research: Provides structural models for studying protein function, molecular mechanisms, and protein interactions.

Methodology:

The method integrates structural models from multiple prediction servers and evaluates them using a consensus scoring scheme based on meta-profile alignments that incorporate predicted secondary structure and sequence conservation information.