Replica-Based Protein Structure Sampling Methods II: Advanced Hybrid Solvent TIGER2hs.

Area of Science:

• Computational chemistry
• Biophysics
• Molecular dynamics simulations

Background:

• Molecular dynamics simulations (MDSs) with modern force fields can predict native protein states.
• Enhanced sampling methods explore protein conformational space and overcome energy barriers.
• Replica exchange molecular dynamics is accurate but computationally expensive.

Purpose of the Study:

• To present an extension of the TIGER2h solvent sampling algorithm, named TIGER2hs.
• To improve the accuracy of hybrid explicit-implicit solvent models in molecular dynamics simulations.
• To better reproduce protein-solvent interactions, including steric effects and reaction field energy.

Main Methods:

• Developed the TIGER2hs algorithm, adding explicit water layers around peptides for energy calculations.
• Maintained dynamics in fully explicit water while using implicit solvent for swap decisions.
• Performed nanoscale molecular dynamics simulations on α-helical (AAQAA)3 and β-hairpin HP7 peptides.

Main Results:

• TIGER2hs better approximates explicit solvent calculations than fully implicit models.
• Explicit water layers reproduce steric effects, solvent polarization, and reaction field energy more accurately.
• A strong conformational dependence of the reaction field energy was observed.

Conclusions:

• TIGER2hs offers a compromise between efficiency and accuracy in molecular dynamics simulations.
• The method approaches the accuracy of full explicit solvent sampling at reduced computational cost.
• This implementation provides a production-ready tool for enhanced protein conformational sampling.