Prediction of Protein-Ligand Binding Pose and Affinity Using the gREST+FEP Method

Area of Science:

• Computational chemistry
• Molecular modeling
• Drug discovery

Background:

• Accurate prediction of protein-ligand binding affinity is crucial for computational chemistry and drug discovery.
• Traditional methods like Free Energy Perturbation (FEP) require high-resolution crystal structures, limiting their applicability.
• A need exists for methods that can predict binding affinity without relying solely on experimental structures.

Purpose of the Study:

• To develop and validate a novel sequential protocol combining generalized replica exchange with solute tempering (gREST) and FEP.
• To enable accurate prediction of protein-ligand binding affinity without requiring high-resolution structural information of the ligand-bound state.
• To enhance the efficiency and scope of *in-silico* drug discovery pipelines.

Main Methods:

• A sequential protocol integrating gREST for pose prediction and FEP for affinity calculation was developed.
• gREST simulations employed high temperatures and flat-bottom restraint potentials to efficiently sample multiple ligand binding poses.
• The FEP method was subsequently applied to the most reliable pose identified by gREST.

Main Results:

• The gREST+FEP protocol demonstrated excellent agreement between calculated and experimental binding affinities for ten ligands binding to FK506 binding proteins (FKBP).
• The method successfully predicted binding affinities without the need for high-resolution X-ray crystallography data.
• The protocol proved effective in overcoming the limitations of traditional FEP methods.

Conclusions:

• The developed gREST+FEP method offers a robust and accurate approach for predicting protein-ligand binding affinities.
• This protocol significantly advances *in-silico* drug discovery by removing the dependency on high-resolution structural data.
• The gREST+FEP method is a valuable tool for accelerating the identification of potential drug candidates.