Molecular dynamics (MD) simulations provide atomistic insight into biological processes but are often limited by poor sampling of rare events and long-timescale transitions. Enhanced sampling methods overcome these limitations by enabling efficient exploration of complex free-energy landscapes. In this talk, I will present enhanced sampling studies of toxin–membrane interactions and the folding of lasso peptides. For mycolactone, a cytotoxic macrolide responsible for Buruli ulcer, simulations with mammalian plasma and endoplasmic reticulum (ER) membrane models reveal a probabilistic preference for the ER, driven by membrane hydration, deformation, and lipid-specific interactions. Isomer-specific simulations further show that cytotoxic mycolactone B stabilizes a closed Sec61 translocon conformation, inhibiting protein translocation. I will also discuss enhanced sampling approaches to characterize lasso peptide folding, identifying reaction coordinates that distinguish pre-lasso from nonthreaded states and demonstrating how sequence optimization and chemical modification can significantly stabilize the pre-lasso ensemble.