Biopolymer structure simulation and optimization via fragment regrowth Monte Carlo

Area of Science:

• Computational biology
• Biophysics
• Statistical mechanics

Background:

• Efficient exploration of biopolymer configuration space is crucial for accurate structure modeling and prediction.
• Existing methods may face computational challenges in simulating complex polymer systems.

Purpose of the Study:

• To introduce a novel Monte Carlo method, fragment regrowth via energy-guided sequential sampling (FRESS), for enhanced chain polymer simulations.
• To improve the efficiency and accuracy of biopolymer structure modeling and prediction.

Main Methods:

• Development and application of the fragment regrowth via energy-guided sequential sampling (FRESS) method.
• Integration of multigrid Monte Carlo concepts within a configurational-bias Monte Carlo framework.
• Testing FRESS on two- and three-dimensional hydrophobic-hydrophilic (HP) protein folding models.

Main Results:

• FRESS achieved all previously known minimum energies for HP models with significantly reduced computation time.
• FRESS identified novel, lower energy configurations for 3D HP models exceeding 80 residues.
• A new extension of the Metropolis Monte Carlo framework was discovered as a byproduct.

Conclusions:

• FRESS offers a computationally efficient and effective approach for biopolymer structure modeling.
• The method demonstrates superior performance in exploring configuration space and identifying low-energy states.
• The findings contribute to advancing computational methods in protein folding and polymer simulations.