Quantitatively Accounting for Protein Reorganization in Computer-Aided Drug Design

Structure-based drug design frequently operates underthe assumptionthat a single holo structure is relevant. However, a large numberof crystallographic examples clearly show that multiple conformationsare possible. In those cases, the protein reorganization free energymust be known to accurately predict binding free energies for ligands.Only then can the energetic preference among these multiple proteinconformations be utilized to design ligands with stronger bindingpotency and selectivity. Here, we present a computational method toquantify these protein reorganization free energies. We test it ontwo retrospective drug design cases, Abl kinase and HSP90, and illustratehow alternative holo conformations can be derisked and lead to largeboosts in affinity. This method will allow computer-aided drug designto better support complex protein targets.

Automated summary

Structure-based drug design often assumes a single relevant holo structure, but multiple conformations have been observed in crystallographic examples. The protein reorganization free energy is crucial for accurately predicting binding free energies and designing ligands with stronger potency and selectivity. In this study, we present a computational method to quantify protein reorganization free energies and demonstrate its effectiveness in two drug design cases. This method will enhance computer-aided drug design for complex protein targets.