PREPRINT (not peer-reviewed). A proposed free-energy method reuses sampled trajectories across tilted states without recalculating molecular energies or instantaneous forces for the reweighting step. Fang and colleagues introduce population-balance tilt sampling: they adjust a linear tilt until equal numbers of samples fall on each side of a restrained window, then use that balance point to estimate the local free-energy slope. Their theory predicts lower leading-order finite-window bias than a mean-balance estimator for a particular cubic asymmetry, at the cost of a modest increase in asymptotic variance. Tests on the Wolfe–Quapp model and gas-phase alanine dipeptide show one- or two-dimensional free-energy reconstruction; the balance-point histories can also reveal local sampling bottlenecks. These are limited model and molecular benchmarks, not a universal accuracy guarantee. Find the paper on ChemRxiv by searching its full title: Population-Balance Tilt Sampling for Gradient-based Free-Energy Reconstruction.
5:38:29 PM · Nov 22, 2026
Sign in to reply and join the conversation.
