Charges for Large Scale Binding Free Energy Calculations with the Linear Interaction Energy Method

Area of Science:

• Computational chemistry
• Molecular modeling
• Drug discovery

Background:

• Linear interaction energy (LIE) method combines molecular dynamics (MD) simulations and linear response theory for binding free energy estimation.
• High-throughput virtual screening requires efficient and automatable methods for assigning atomic charges in molecular mechanics force fields like OPLS-AA.

Purpose of the Study:

• To evaluate various ab initio and semiempirical charge methods for automated LIE calculations.
• To identify optimal charge assignment strategies for high-throughput virtual screening of drug candidates.

Main Methods:

• Assessed nine ab initio and semiempirical charge methods, including solvent-induced polarization estimates.
• Calculated relative binding free energies for ten HIV-1 reverse transcriptase inhibitors using 23 charge variants via LIE.
• Performed over 800 ns of MD simulations to validate charge models.

Main Results:

• The LIE method demonstrated excellent binding free energy estimates across multiple charge assignment strategies.
• Semiempirically derived CM1A charges proved to be a fast and reliable alternative for automated LIE calculations.
• Conclusions are applicable to other free energy calculation methods like FEP and TI.

Conclusions:

• Automated charge assignment is crucial for high-throughput virtual screening using LIE.
• CM1A charges present a viable and efficient option for LIE-based virtual screening with OPLS-AA.
• The findings support the broader applicability of these charge models in computational drug discovery.