Description and assessment of a model for GSK-3beta database virtual screening

Area of Science:

• Medicinal Chemistry
• Computational Biology
• Drug Discovery

Background:

• Glycogen synthase kinase 3 beta (GSK-3beta) is a key target in various diseases.
• Accurate in silico models are crucial for efficient virtual screening of potential drug candidates.
• Prioritizing compounds for synthesis and experimental testing saves significant resources.

Purpose of the Study:

• To develop and validate a robust in silico methodology for generating reliable GSK-3beta protein models.
• To identify optimal molecular docking and scoring strategies for predicting inhibitor activity.
• To facilitate the prioritization of compounds for experimental evaluation in drug discovery programs.

Main Methods:

• Generation of multiple in silico GSK-3beta protein models from X-ray crystallographic data.
• Flexible molecular docking of 42 known GSK-3beta inhibitors into each protein model.
• Re-scoring of docked poses using various scoring functions and comparison with experimental biological activities.

Main Results:

• Different protein models and scoring functions yielded varying rankings of inhibitor efficacy.
• A specific GSK-3beta model combined with two scoring functions demonstrated the highest correlation with known biological activities.
• This validated approach effectively prioritizes potential inhibitors for further investigation.

Conclusions:

• The developed methodology offers an accessible and accurate strategy for creating reliable models for virtual database screening.
• This approach can significantly streamline the drug discovery process by efficiently identifying promising GSK-3beta inhibitors.
• The findings support the use of validated in silico models in prioritizing chemical synthesis and experimental evaluations.