Navigating the Unknown: Discovering Minimum Free Energy Pathways without Predefined End States

Determining minimum free energy pathways (MFEPs) for protein conformational changes is essential for molecular-level mechanistic understanding. While many robust path-search algorithms have existed, most require end point conformations derived from experimental structures, creating a dependency on structural biology data that restricts their general applicability. To overcome this limitation, we present a new path-search algorithm based on local sampling. The process can initiate from any single state, while the search direction is automatically optimized without the information on other states. We demonstrate the effectiveness of this algorithm in several model systems by comparing with experimental data and conventional molecular dynamics simulations. This approach expands the methodological toolkit for investigating functional conformational transitions, particularly when experimental end point structures are unavailable.