Derisking Affinity Optimization for Macrocycles and Cyclic Peptides: High-Precision Free Energy Simulations across...

Area of Science:

• Computational chemistry and drug discovery
• Medicinal chemistry
• Pharmacology

Background:

• Macrocycles and cyclic peptides are promising for targeting difficult biological targets.
• Their complex structures and synthesis present challenges in drug discovery.
• Physics-based computational methods are crucial for prioritizing macrocyclic drug candidates.

Purpose of the Study:

• To validate the Free Energy Perturbation (FEP+) framework for predicting binding affinities of macrocyclic and cyclic peptide inhibitors.
• To assess the accuracy and reliability of FEP+ across diverse therapeutic targets.
• To demonstrate the utility of FEP+ in accelerating the drug discovery process for beyond-rule-of-five (bRo5) therapeutics.

Main Methods:

• Retrospective validation of the FEP+ computational framework.
• Application to five diverse macrocyclic and cyclic peptide inhibitor series (KRAS, PCSK9, MCL-1, JAK2, Cyclin A/B).
• Analysis of over 230 unique peptidic and nonpeptidic analogues.

Main Results:

• Achieved robust predictive accuracy with a global pairwise RMSE_ΔΔG of 1.06 kcal/mol.
• Demonstrated reliable intraseries rank-ordering of drug candidates.
• Showcased robust absolute accuracy across a >10 kcal/mol experimental dynamic range (>7 orders of magnitude in binding affinity).

Conclusions:

• FEP+ is an effective computational method for derisking macrocyclic drug discovery.
• The framework provides critical insights into ligand preorganization, hydration, and binding energetics.
• FEP+ accelerates the identification of clinically viable beyond-rule-of-five (bRo5) therapeutics.