Improved accuracy of hybrid atomistic/coarse-grained simulations using reparametrised interactions.

Area of Science:

• Computational chemistry and biophysics.
• Development of advanced simulation methodologies.

Background:

• Coarse-graining (CG) in molecular dynamics reduces computational cost but sacrifices information.
• Hybrid atomistic (AT)/CG methods offer a balance, but previous AT-CG interactions caused artificial increases in protein intramolecular hydrogen bonds, reducing flexibility.
• Existing AT-CG parameters showed unfavorable solvation for polar solutes and favorable for apolar ones.

Purpose of the Study:

• To reparametrize atomistic/coarse-grained (AT/CG) interactions to improve accuracy in hybrid simulations.
• To enhance the reproduction of protein structural properties and flexibility in CG water.
• To eliminate the need for an explicit atomistic solvent layer.

Main Methods:

• Reparametrization of AT-CG interactions based on solvation free energies of alkanes and side-chain analogues in CG water.
• Testing new AT-CG parameters in hybrid simulations of four proteins using CG water.
• Comparison of results from hybrid simulations with fully atomistic simulations and experimental data.

Main Results:

• The reparametrized AT-CG interactions significantly reduced the artificial increase in intramolecular hydrogen bonds.
• Hybrid simulations with new parameters showed improved reproduction of protein structural properties and flexibility.
• The method successfully maintained good mixing of AT and CG water.

Conclusions:

• The refined AT-CG parameters provide a more accurate and flexible representation of proteins in coarse-grained simulations.
• This approach enhances the reliability of hybrid simulations for studying protein dynamics and structure.
• The developed method eliminates the necessity of an atomistic solvent layer, further increasing computational efficiency.