MuD

MuD predicts the functional impact of missense mutations on proteins by integrating sequence-derived and structural features to classify amino acid substitutions as likely neutral or deleterious.

Key Features:

• Random Forests-Based Classifier: Uses a Random Forests machine-learning classifier to evaluate the potential effects of missense mutations.
• Sequence and Structural Features: Integrates diverse sequence-derived and structural features for comprehensive prediction of mutation impact.
• User-Specific Enhancements: Incorporates user-reported protein-specific structural and functional information at runtime to tailor predictions.
• Reliability Scoring: Assigns a reliability score to each prediction to assist prioritization, particularly for proteins with available 3D structures.
• Automatic Mode Performance: In automatic mode, performs comparably to other existing tools.

Scientific Applications:

• Variant interpretation: Classifies missense variants as likely neutral or deleterious to aid interpretation of protein-coding mutations.
• Mutation prioritization: Prioritizes amino acid substitutions for experimental follow-up using reliability scores and structural information.
• Disease mechanism research: Supports genomics and proteomics studies investigating the molecular consequences of substitutions in disease-associated proteins.

Methodology:

Random Forests machine-learning classifier using sequence-derived and structural features, accepting user-reported protein-specific structural and functional information at runtime, producing a per-prediction reliability score and supporting an automatic mode.