Binding-site assessment by virtual fragment screening.

Area of Science:

• Computational chemistry
• Drug discovery
• Chemical genomics

Background:

• Accurate prediction of protein druggability is crucial for advancing chemical genomics and drug discovery.
• Current methods for assessing druggability can be enhanced by novel computational approaches.

Purpose of the Study:

• To develop and validate a novel computational method for quantitatively assessing protein druggability.
• To evaluate the method's ability to distinguish between druggable and non-druggable protein targets.
• To explore the method's utility in identifying druggable conformations and guiding lead optimization.

Main Methods:

• Computationally screening a library of approximately 11,000 fragment-like compounds against protein binding sites.
• Docking fragments and calculating a computational hit rate based on a score cutoff.
• Large-scale evaluation on 152 binding sites across four datasets.
• Assessing sensitivity to different receptor conformations, including flexible protein-protein interaction sites.

Main Results:

• Computed hit rates showed a significant correlation with experimentally measured hit rates from NMR-based fragment screening.
• The in silico fragment screening method successfully differentiated between known druggable and non-druggable targets (enzymes and protein-protein interaction sites).
• The method's performance was evaluated across various receptor conformations, demonstrating robustness.

Conclusions:

• The developed computational fragment screening method provides a reliable approach for assessing protein druggability.
• This method can aid in identifying potential drug targets and optimizing lead compounds in drug discovery.
• The approach offers insights into druggable binding site conformations and strategies for fragment-based drug design.