Assessment of a novel scoring method based on solvent accessible surface area descriptors

Area of Science:

• Computational chemistry
• Structural biology
• Drug discovery

Background:

• Virtual screening (VS) is crucial for identifying drug candidates.
• Existing VS methods like GOLD and Glide have limitations.
• Novel scoring algorithms are needed to improve efficiency and accuracy.

Purpose of the Study:

• To develop and evaluate a novel scoring algorithm based on solvent accessible surface area (SASA) descriptors.
• To compare the database enrichment potential of the SASA-based algorithm against established VS methods (GOLD and Glide).
• To assess the robustness and applicability of the SASA descriptors across diverse protein targets.

Main Methods:

• Utilized GOLD for structure-based VS to generate protein-ligand docking poses.
• Applied a protein-ligand interaction fingerprint metric for postprocessing docking poses.
• Computed SASA descriptors for ligands and proteins in bound/unbound states.
• Developed a Bayesian model using SASA descriptors to score screening databases.

Main Results:

• The SASA-based scoring algorithm demonstrated early database enrichment comparable or superior to GOLD and Glide.
• The algorithm exhibited robustness across various protein target classes.
• Satisfactory early enrichment was achieved for a wide range of targets.

Conclusions:

• Novel topological descriptors based on SASA are a valuable in silico tool for hit identification.
• The SASA-based scoring algorithm offers a promising alternative to existing VS methods.
• This approach can enhance the efficiency of drug discovery pipelines.