Umbrella Sampling Workflows for Fast-Converging PMF Calculations without Artificial WHAM Constraints

Journal article

Umbrella sampling is a powerful tool to investigate the underlying free-energy landscapes that govern the permeation properties of small molecules through complex lipid bilayers. However, obtaining converged potentials of mean force (PMFs) remains an issue due to sampling limitations, and as a result, enhanced sampling approaches are constantly being developed. Here, we benchmark umbrella sampling workflows and test different options for window generation, sampling, and statistical estimation of PMFs against each other to arrive at recommendations to improve PMF convergence speeds and minimize required computational resources. We also report large errors introduced into PMFs by enforcing their symmetry and/or periodicity when sampling is insufficient, leading us to caution against the use of these constraints in future works.

Authors: Feddersen, B; Biggin, PC

• Publication status: Published
• Peer review status: Peer reviewed
• Publisher: American Chemical Society
• Journal: Journal of Chemical Theory and Computation
• Volume: 22
• Issue: 10
• Pages: 5299-5310
• Publication date: 2026-05-06
• Acceptance date: 2026-04-28
• DOI: 10.1021/acs.jctc.6c00246
• EISSN: 1549-9626
• ISSN: 1549-9618
• Language: English
• Keywords: Sampling (signal processing); Workflow; Benchmark (surveying); Umbrella sampling; Convergence (economics); Window (computing); Importance sampling