Modeling DNA Flexibility: Comparison of Force Fields from Atomistic to Multiscale Levels

Area of Science:

• Computational Chemistry
• Molecular Dynamics
• Biophysics

Background:

• Accurate force fields (FFs) are essential for reliable computer simulations in molecular biology.
• Evaluating FF performance is critical for understanding DNA dynamics and flexibility.

Purpose of the Study:

• To assess the performance of four major force fields (CHARMM and AMBER families) in simulating DNA dynamics.
• To compare atomistic and coarse-grained (CG) simulation results for DNA flexibility and stability.
• To evaluate the ability of FFs to parameterize CG DNA models.

Main Methods:

• Multi-microsecond molecular dynamics simulations of a 40-base-pair DNA fragment.
• Analysis of DNA flexibility, dynamics, base-pair/step parameters, and persistence length.
• Structure-based bottom-up coarse-graining for CG DNA model parameterization.

Main Results:

• CHARMM36 failed to maintain DNA structural stability over microsecond timescales.
• AMBER force fields (parmbsc0, parmbsc1) showed good agreement with experimental data, with parmbsc0 exhibiting bias for A/B DNA forms.
• CHARMM27 provided stable trajectories and the best overall fit to experimental DNA flexibility parameters.

Conclusions:

• CHARMM27 is the most suitable force field among those tested for atomistic and CG DNA simulations.
• Force field choice significantly impacts the accuracy of DNA flexibility and dynamics predictions.
• Accurate FF parametrization is key for advancing predictive power in molecular simulations.