SIENA

SIENA generates binding site ensembles and identifies alternative binding site conformations from Protein Data Bank structures to characterize protein structural flexibility for molecular recognition and enzymatic function.

Key Features:

• Automated assembly and preprocessing: Generates binding site ensembles from an arbitrarily defined binding site in a single protein structure and handles structural artifacts and inconsistent PDB annotations.
• Identification of alternative conformations: Searches sequentially similar proteins to retrieve alternative binding site conformations using an indexed database and perfect k-mer matches against the Protein Data Bank.
• Alignment algorithm - ASCONA: Aligns multiple binding site conformations, detects flexible backbone deviations, manages multiple binding site matches within single structures, and operates at an average computation time of 4 milliseconds per target.
• Interaction-based selection: Selects binding site conformations based on interaction patterns to cover known ligand-binding geometries and to accommodate both known and previously unconsidered ligand molecules.
• Ensemble generation efficiency: Generates complete binding site ensembles from the entire Protein Data Bank in a few seconds.

Scientific Applications:

• Modeling protein flexibility: Provides ensembles that represent alternative conformations for studying backbone and side-chain variability in binding sites.
• Structure-based virtual screening: Supplies diverse binding site conformations to improve ligand docking and virtual screening campaigns.
• Protein–ligand interaction analysis: Enables analysis of interaction patterns across multiple binding site conformations to identify conserved and variable contacts.
• Point mutation studies: Facilitates assessment of how mutations affect binding site conformations and potential ligand binding.
• Enzymatic function exploration: Supports exploration of enzymatic mechanisms under different conformational states by providing alternative binding site geometries.

Methodology:

Starts from an arbitrarily defined binding site in a single structure, assembles and preprocesses ensembles, searches sequentially similar proteins using an indexed database with perfect k-mer matches to retrieve PDB entries, aligns conformations with the ASCONA algorithm that detects flexible backbone deviations and handles multiple matches, and selects conformations via an interaction-based algorithm to cover ligand-binding geometries; reported timings are ~4 ms per target for ASCONA and ensemble generation from the whole PDB in a few seconds.