Adaptive Biasing Potential Method: Difference between revisions

From PMFLib Wiki
Jump to navigation Jump to search
No edit summary
No edit summary
 
Line 2: Line 2:
----
----
=Introduction=
=Introduction=
The Adaptive Biasing Potential (ABP) method implemented in PMFLib follows the mollified density-of-states formulation [1]. During the simulation, the method accumulates a smoothed population of the selected collective variables and uses it to construct an adaptive biasing force. The directly obtained free-energy estimate is therefore the mollified free energy, not the exact free energy surface. To recover the final free energy, the mollification error must be removed by post-processing, typically via deconvolution. In PMFLib, this correction and reconstruction of the free-energy profile are performed with the [[abp-energy]] utility.
The Adaptive Biasing Potential (ABP) method implemented in PMFLib follows the mollified density-of-states formulation [1]. During the simulation, ABP uses the selected collective variables to construct a discretised, mollified population, which is then used to compute the mollified free energy and the corresponding adaptive biasing force. The directly obtained free-energy estimate is therefore the mollified free energy, not the exact free-energy surface. To recover the final free energy, the mollification error must be removed by post-processing, typically via deconvolution. In PMFLib, this correction and reconstruction of the free-energy profile are performed with the [[abp-energy]] utility. ABP also supports the Multiple-Walker Approach, in which several simulations contribute to a shared accumulator through the MWA server, accelerating the construction of the mollified population and the convergence of the adaptive bias.


=Documentation=
=Documentation=

Latest revision as of 13:27, 20 June 2026

Navigation: Documentation / Methods / Adaptive Biasing Potential Method


Introduction

The Adaptive Biasing Potential (ABP) method implemented in PMFLib follows the mollified density-of-states formulation [1]. During the simulation, ABP uses the selected collective variables to construct a discretised, mollified population, which is then used to compute the mollified free energy and the corresponding adaptive biasing force. The directly obtained free-energy estimate is therefore the mollified free energy, not the exact free-energy surface. To recover the final free energy, the mollification error must be removed by post-processing, typically via deconvolution. In PMFLib, this correction and reconstruction of the free-energy profile are performed with the abp-energy utility. ABP also supports the Multiple-Walker Approach, in which several simulations contribute to a shared accumulator through the MWA server, accelerating the construction of the mollified population and the convergence of the adaptive bias.

Documentation


References

(1) Dickson, B. M.; Legoll, F.; Lelièvre, T.; Stoltz, G.; Fleurat-Lessard, P. Free Energy Calculations: An Efficient Adaptive Biasing Potential Method. J. Phys. Chem. B 2010, 114 (17), 5823–5830. https://doi.org/10.1021/jp100926h.