Development and Validation of the Quantum Mechanical Bespoke Protein Force Field

Area of Science:

• Computational chemistry and structural biology
• Development of molecular mechanics force fields

Background:

• Traditional molecular mechanics force fields for proteins often neglect system-specific polarization by fitting to small molecule properties.
• This limitation can affect the accuracy of simulations for biological macromolecules.

Purpose of the Study:

• To introduce a complete protein force field compatible with the quantum mechanical bespoke (QUBE) force field.
• To derive nonbonded parameters directly from the electron density of specific proteins.
• To rederive backbone and sidechain torsional parameters using quantum mechanical dihedral scans.

Main Methods:

• Derivation of nonbonded parameters from protein-specific electron density.
• Fitting of torsional parameters to quantum mechanical dihedral scans.
• Validation through molecular dynamics simulations of peptides and proteins, comparing conformational preferences with experimental data (e.g., NMR J couplings).

Main Results:

• Accurate backbone and sidechain conformations were obtained in dipeptide simulations, with NMR J coupling errors comparable to the OPLS force field.
• Secondary structures were generally retained in simulations of five folded proteins.
• NMR J coupling errors in folded protein simulations were similar to standard transferable force fields, though some regional structural deviations were noted.

Conclusions:

• The developed protein force field, utilizing system-specific nonbonded parameters, shows promise for accurate molecular dynamics simulations.
• This approach offers a viable path for next-generation simulations of biological molecules, improving upon traditional methods.
• Further development is warranted to fully leverage system-specific parameters for enhanced simulation accuracy.