Accelerated Discovery of Carbamate Cbl-b Inhibitors Using Generative AI Models and Structure-Based Drug Design

Area of Science:

• Immunology
• Medicinal Chemistry
• Artificial Intelligence in Drug Discovery

Background:

• Casitas B-lymphoma proto-oncogene-b (Cbl-b) is a key E3 ligase regulating T cell, NK cell, and B cell activation.
• Cbl-b acts as a negative regulator, making it a target for modulating immune responses.

Purpose of the Study:

• To discover novel inhibitors of Cbl-b using a combination of generative AI and structure-based drug design.
• To accelerate the drug discovery process for Cbl-b inhibitors.

Main Methods:

• Integration of the REINVENT generative AI engine with medicinal chemistry structure-based design.
• Iterative in silico structure-based drug design within the Design-Make-Test-Analyze cycle.
• Utilized physics-based affinity prediction and machine learning DMPK models to guide design.

Main Results:

• Successfully discovered a potent series of carbamate Cbl-b inhibitors.
• Demonstrated accelerated discovery through the integrated AI and structure-based design approach.
• Validated the effectiveness of in silico predictive models in guiding synthesis selection.

Conclusions:

• The combined generative AI and structure-based design approach significantly accelerates the discovery of novel drug candidates.
• This strategy efficiently identified potent Cbl-b inhibitors, highlighting its potential for future drug discovery efforts.
• Optimized design phase improves efficiency and success rate in identifying targeted molecular inhibitors.