Pocketome

Pocketome catalogs experimentally solved conformational ensembles of druggable protein binding sites to characterize binding site plasticity and support analysis of protein-ligand interactions for drug discovery.

Key Features:

• Curated ensembles: Approximately one thousand conformational ensembles derived from the Protein Data Bank (PDB) and the Uniprot Knowledgebase.
• Grouping by site: Ensembles are grouped by their location within proteins while maintaining consistent chain and cofactor compositions.
• Multiple pockets per ensemble: Each ensemble can include multiple pockets to capture different conformational states of binding sites.
• Pairwise classification: Pockets are classified according to pairwise structural similarity and compatibility with different ligands.
• Residue annotation: Identification of key residues involved in binding interactions is provided for each pocket.
• Binding compatibility matrices: Matrices are constructed to represent compatibility between pockets and ligands.
• Automated updates and curation: Core data are updated through automated integration of new PDB and Uniprot Knowledgebase releases and supplemented by manually curated entries based on seed ligand locations.
• Analytical functions: Functionality includes searching specific sites, analyzing conformational clusters, identifying important residues, and generating binding compatibility matrices.
• Visualization integration: Supports interactive visualization of ensembles via the ActiveICM web browser plugin.

Scientific Applications:

• Multi-conformational docking models: Construction of docking models that incorporate multiple binding-site conformations.
• 3D activity models: Development of three-dimensional activity models that account for site flexibility.
• Cross-docking experiment design: Design and interpretation of cross-docking experiments to evaluate ligand compatibility across conformations.
• Virtual screening benchmarks: Creation of virtual ligand screening benchmarks that reflect binding-site plasticity.
• Ligand affinity and specificity prediction: Support for predicting ligand affinity and specificity using pocket classifications and compatibility data.
• Understanding dynamic interactions: Analysis of dynamic protein-ligand interactions informed by experimentally solved conformational ensembles.

Methodology:

Conformational ensembles are derived from experimentally solved structures in the Protein Data Bank (PDB) and the Uniprot Knowledgebase, grouped by site with consistent chain and cofactor composition, classified by pairwise structural similarity and ligand compatibility, assembled into binding compatibility matrices, and maintained via automated integration of new PDB and Uniprot Knowledgebase releases supplemented by manual curation based on seed ligand locations.