Treating Polarization Effects in Charged and Polar Bio-Molecules Through Variable Electrostatic Parameters.

Area of Science:

• Computational chemistry
• Molecular modeling
• Drug discovery

Background:

• Polarization effects are crucial in systems with charges and hydrogen bonds.
• Current polarizable force fields face limitations in computational cost and transferability.
• Existing methods struggle to balance accuracy and efficiency in modeling polarization.

Purpose of the Study:

• To develop more efficient and accurate methods for incorporating polarization effects into molecular simulations.
• To investigate the combined impact of polarization and entropy on predicting drug-like molecule properties.
• To explore the integration of quantum mechanics (QM) and machine learning (ML) for enhanced polarization modeling.

Main Methods:

• Utilizing QM-based and ML-assisted models to capture polarization effects.
• Employing linear scaling fragmentation-based QM calculations for reduced computational expense.
• Performing all-atom molecular dynamics simulations (explicit and implicit) to assess partition coefficients.

Main Results:

• Demonstrated efficient treatment of polarization effects by combining QM accuracy with MM/ML efficiency.
• Reduced computational costs through variable electrostatic parameters (charge, dipole, dielectric constant).
• Quantified the influence of polarization and entropy on the prediction of partition coefficients for druglike molecules.

Conclusions:

• QM/ML approaches offer a promising avenue for accurate and efficient polarization modeling.
• Addressing computational cost and transferability are key challenges in polarizable force field development.
• Further research is needed to refine these methods for broader applications in molecular simulations and drug design.