GSP4PDB

GSP4PDB represents protein-ligand structural patterns as graph models and enables their search and analysis across the Protein Data Bank (PDB) to characterize protein-ligand interactions and inform protein engineering and biotechnology.

Key Features:

• Graph-based Structural Patterns (GSPs): Abstracts protein-ligand complexes into graph models where nodes correspond to components such as amino acids and ligands and edges denote structural relationships, including distances between ligands and amino acids.
• Relational database search of PDB: Allows defined GSPs to be matched against a relational database derived from Protein Data Bank (PDB) entries, returning matches represented as graph patterns.
• Filtering and data export: Provides filters to refine search results and exports related structural information for external post-processing and analysis.

Scientific Applications:

• Protein engineering: Identify and compare recurring structural interaction patterns to guide protein design and modification.
• Drug design: Characterize protein-ligand interactions and recurrent binding motifs across PDB entries to inform lead optimization.
• Molecular and structural biology: Discover and annotate interaction patterns from PDB structural data to infer functional relationships.

Methodology:

Protein-ligand complexes are abstracted into Graph-based Structural Patterns (GSPs) where nodes represent amino acids and ligands and edges denote structural relationships, including distances between ligands and amino acids; these GSPs are matched against a relational database derived from PDB entries, and results can be filtered and exported for external post-processing.