Using generative AI to transform peptide hits into small molecule leads

Area of Science:

• Computational chemistry
• Drug discovery
• Artificial intelligence in medicine

Background:

• Abundant structural data exists for peptide-protein interactions, offering a foundation for small molecule inhibitor design.
• Improving pharmacokinetic properties of peptide-based therapeutics is a key challenge in drug development.

Purpose of the Study:

• To highlight the potential of AI-enabled tools in transforming peptide hits into small molecule leads.
• To outline an AI-driven workflow for structure-based drug design, from peptide to small molecule.

Main Methods:

• Utilizing AI for target protein structure prediction.
• Employing generative models for de novo peptide binder design.
• Leveraging diffusion models for novel small molecule scaffold generation.
• Applying deep learning for binding affinity prediction and candidate triage.

Main Results:

• AI tools can potentially automate and expedite multiple stages of the drug design pipeline.
• Integration of AI can enhance the efficiency of converting peptide binders into viable small molecule drug candidates.
• AI facilitates rapid screening and selection of promising small molecules based on predicted binding affinity.

Conclusions:

• AI presents a transformative approach to structure-based drug design, significantly accelerating the journey from peptide binders to small molecule leads.
• The described AI-enabled workflow offers a comprehensive strategy for optimizing drug discovery processes.
• Future applications of AI in medicinal chemistry hold promise for developing improved therapeutics with enhanced pharmacokinetic profiles.