I-MUTANT

I-MUTANT predicts changes in protein stability upon single point mutations using support vector machine (SVM) models with protein sequence or structure input.

Key Features:

• SVM algorithm: Employs support vector machine (SVM) models to predict stability changes.
• Input types: Accepts protein amino acid sequence or three-dimensional structure as input to enable predictions with or without atomic-resolution structures.
• Training and validation: Trained and validated on experimental thermodynamic data from the ProTherm database.
• Classification mode: Predicts the sign of stability change (increase or decrease) with approximately 80% accuracy using structural information and 77% accuracy using sequence data.
• Regression estimation: Estimates DeltaDeltaG (ΔΔG) values with correlations of 0.71 (standard error 1.30 kcal/mol) for structure-based predictions and 0.62 (standard error 1.45 kcal/mol) for sequence-based predictions.

Scientific Applications:

• Protein design: Predicts effects of single point mutations to guide protein engineering and stability optimization.
• Disease mutation analysis: Interprets potential impacts of missense mutations on protein stability in disease contexts.
• Enzyme optimization: Supports selection of stabilizing mutations for industrial enzyme improvement.
• Drug discovery: Assesses how mutations may alter stability of therapeutic targets relevant to drug development.

Methodology:

Uses SVM-based algorithms trained and validated on ProTherm to analyze protein sequence or structure and provide classification of stability sign and regression estimates of DeltaDeltaG (ΔΔG).