Enhanced Sampling with Suboptimal Collective Variables: Reconciling Accuracy and Convergence Speed

Area of Science:

• Computational Chemistry
• Molecular Dynamics
• Biophysics

Background:

• Accurate free energy calculations are crucial for understanding molecular mechanisms.
• Rare events in molecular systems often require enhanced sampling techniques.
• Optimal selection of collective variables (CVs) can be challenging for enhanced sampling.

Purpose of the Study:

• To develop an enhanced sampling algorithm for converged free energy landscapes of molecular rare events.
• To improve transition rates between metastable states and accelerate free energy estimation.
• To provide a robust method for complex biomolecular systems, irrespective of CV choice.

Main Methods:

• Combining on-the-fly probability enhanced sampling (OPES) with its exploratory variant (OPES Explore).
• Sampling a time-dependent target distribution to enhance exploration.
• Applying the algorithm to diverse systems: Wolfe-Quapp potential, trypsin-benzamidine binding, and chignolin protein folding.

Main Results:

• Achieved converged free energy landscapes for molecular rare events.
• Demonstrated accelerated convergence of free energy estimates.
• Validated the algorithm's robustness across different systems and suboptimal CVs.

Conclusions:

• The enhanced sampling algorithm offers accurate free energy calculations at a reduced computational cost.
• The method is robust to the choice of collective variables (CVs).
• This approach holds significant promise for simulating complex biomolecular systems.