ConQuass

ConQuass assesses the quality of protein structural models by evaluating consistency between predicted structures and evolutionary conservation patterns to identify inaccuracies and support model validation.

Key Features:

• Conservation-Based Quality Assessment: Correlates model structural features with residue-level evolutionary conservation, leveraging the principle that conserved residues tend to occupy structural cores while variable residues map to surfaces.
• Identification of Faulty Models: Detects problematic structural models by analyzing discrepancies between predicted structures and expected conservation patterns.
• Performance Correlation with Native Structures: Assigned scores for server models in CASP8 correlate strongly with model similarity to native structures.
• Complementary Methodology: Performs comparably to and complements other single-structure quality assessment methods in CASP8 when conservation information is reliable, and is applicable where homolog-derived structural information is not used.

Scientific Applications:

• Protein structure prediction and validation: Assesses and ranks predicted models to support model selection and refinement.
• Structural biology: Provides conservation-informed quality metrics to interpret experimental or predicted structures.
• Drug design: Helps ensure structural accuracy of target models used in structure-based drug design.
• Molecular dynamics simulations: Identifies regions of potential structural inaccuracy that could affect simulation setup and interpretation.

Methodology:

Perl-based implementation using Perl and R scripts to evaluate alignment between predicted model structures and evolutionary conservation patterns.