Rapid equilibrium sampling initiated from nonequilibrium data.

Area of Science:

• Computational Biology
• Biophysics
• Molecular Dynamics

Background:

• Simulating biomolecular conformational dynamics is challenging due to complex free energy landscapes and metastable states.
• Generalized ensemble (GE) algorithms aid state transitions via temperature space walks, but struggle with entropic barriers dominant in many biological systems.

Purpose of the Study:

• To introduce a novel, efficient algorithm, the adaptive seeding method (ASM), for enhanced conformational sampling of biomolecules.
• To address limitations of existing methods in overcoming entropic barriers during simulations.

Main Methods:

• The ASM employs nonequilibrium GE simulations to identify metastable states.
• Short, constant-temperature simulations are seeded from identified states to determine equilibrium populations.
• The method leverages GE's broad sampling while efficiently crossing entropic barriers at low temperatures.

Main Results:

• The ASM demonstrates efficiency in crossing entropic barriers, particularly for systems like short RNA hairpins.
• Only local equilibrium is required, enabling the use of very short seeding simulations.
• The algorithm can recover equilibrium properties from unconverged datasets.

Conclusions:

• The adaptive seeding method (ASM) provides an efficient and versatile approach for conformational sampling in biomolecular simulations.
• ASM is well-suited for modern computing clusters and can improve the accuracy of molecular dynamics studies.