Active Learning FEP Using 3D-QSAR for Prioritizing Bioisosteres in Medicinal Chemistry

Area of Science:

• Medicinal Chemistry
• Computational Chemistry
• Drug Discovery

Background:

• Bioisostere replacement is crucial for optimizing drug candidate potency and selectivity.
• Efficiently selecting bioisosteres is key to successful drug discovery projects.
• A large pool of potential bioisosteres requires effective prioritization methods.

Purpose of the Study:

• To develop and demonstrate an active learning workflow for prioritizing bioisosteric replacements.
• To combine 3D-QSAR and relative binding free energy calculations for enhanced prioritization.
• To accelerate the identification of potent and selective drug candidates.

Main Methods:

• Integration of 3D-quantitative structure-activity relationship (3D-QSAR) models.
• Application of relative binding free energy (RBFE) calculations.
• Development of an active learning workflow to iteratively prioritize molecules.

Main Results:

• The workflow successfully prioritized bioisosteric replacements from a large pool.
• Demonstrated rapid identification of strongest-binding bioisosteres on a human aldose reductase test case.
• Achieved efficient prioritization with a modest computational cost.

Conclusions:

• The combined 3D-QSAR and RBFE active learning workflow is effective for bioisostere prioritization in drug discovery.
• This computational approach significantly enhances the efficiency of lead optimization.
• The method offers a valuable tool for accelerating the discovery of new therapeutics.