PairMap: An Intermediate Insertion Approach for Improving the Accuracy of Relative Free Energy Perturbation...

Area of Science:

• Computational chemistry
• Drug discovery
• Molecular modeling

Background:

• Accurate prediction of binding free energy differences is vital for cost-effective drug discovery.
• Relative binding free energy perturbation (RBFEP) calculations struggle with large topological changes, causing errors and convergence issues.

Purpose of the Study:

• To develop a novel approach, PairMap, for introducing effective intermediates in RBFEP calculations of complex molecular transformations.
• To enhance the accuracy and reduce the computational cost of RBFEP calculations for drug discovery.

Main Methods:

• PairMap exhaustively generates intermediates and determines optimal conversion paths.
• It incorporates thermodynamic cycles into the perturbation map for improved accuracy.
• The method comprehensively considers intermediates, surpassing existing simple approaches.

Main Results:

• PairMap achieved a mean absolute error of 0.93 kcal/mol on a benchmark set, outperforming the conventional Flare FEP method (1.70 kcal/mol).
• In scaffold hopping experiments for the PDE5a target, PairMap yielded more accurate predictions than ABFEP calculations, aligning better with experimental data.
• PairMap effectively handles congeneric series, resolving complex links with minimal additions.

Conclusions:

• PairMap overcomes limitations of existing methods, enabling RBFEP calculations for complex transformations.
• This advancement streamlines lead optimization in drug discovery by improving prediction accuracy and efficiency.