Achieving Sustained Conformational Exploration with Continually Adaptive Local Memory Sampling
Year: 2026
Type: preprint
Abstract: Enhanced sampling techniques aim to expedite the exploration of molecular conformations and their associated transitions. The extended adaptive biasing force method (eABF) is an enhanced sampling technique that applies a biasing force to targeted collective variables (CV), which serves to overcome energy barriers that would normally delay conformational transitions. With eABF, the biasing force is adaptively constructed to nullify the average force acting along the CV, effectively removing barriers to promote accelerated transitions. However, the eABF biasing force can become unresponsive to CV-coupled environmental dynamics, leading to hindered CV sampling. Therefore, in order to promote persistent CV transitions, we developed the continually adaptive local memory (CALM) sampling method, designed to sustain adaptation of the eABF biasing force. This is achieved by maintaining a limited sampling memory, which allows for CV moves to continually synchronize with coupled environmental dynamics. Herein, we show CALM effectively accelerates conformational sampling of alanine dipeptide and deca-alanine in vacuum and also deca-alanine and the mini-protein chignolin in explicit water. As a result of maintaining biasing force adaptability, we illustrate how CALM promotes sustained CV transitions via access to hidden conformational pathways that are not accessible via eABF.
Source: ChemRxiv
Authors: Steven L. Austin, H. Lee Woodcock
Institutions: New College of Florida, University of South Florida
Language: English
Subfield: Molecular Biology
Domain: Life Sciences
Open Access status: green
OpenAlex ID: W7213298196