GASS-Metal

GASS-Metal predicts metal-binding sites on protein structures using a parallel genetic algorithm to identify structural similarities with curated templates from M-CSA and MetalPDB for functional annotation.

Key Features:

• Genetic algorithm-based approach: Employs a parallel genetic algorithm to identify candidate metal-binding sites by searching for structural similarities between target proteins and template sites.
• Template databases: Uses curated templates from M-CSA (Metal-Binding Sites in Proteins) and MetalPDB to guide structural comparisons.
• Robust validation: Validation was performed using homologous proteins and conservative mutations of residues to assess prediction robustness.
• Performance metrics: Reported performance includes a Matthews Correlation Coefficient (MCC) up to 0.57 and correct identification of up to 96.1% of metal-binding sites.

Scientific Applications:

• Structural biology: Predicts metal-binding sites to support interpretation of protein function and metalloprotein characterization.
• Bioinformatics annotation: Provides metal-binding site annotations for computational analyses of proteins and metalloproteomes.
• Experimental design: Guides design of mutagenesis and biochemical experiments to test metal–protein interactions and residue roles.
• Drug discovery: Identifies metal-binding sites relevant to the development of metallo-drugs or inhibitors targeting metalloproteins.

Methodology:

Employs a parallel genetic algorithm to search for structural similarities between candidate sites on target proteins and curated templates from M-CSA and MetalPDB, with validation using homologous proteins and conservative residue mutations.