Dynamical Fingerprints of Rare-Event Transitions from Transition Corridors

Abstract

Rare transitions are central in science, yet their mechanisms are obscured by the high dimensionality of complex dynamics, a challenge particularly pronounced in biomolecules where rare conformational transitions underlie functional processes. A transition-conditioned dynamical fingerprint is introduced for a rare \(A \mathbin{\text{--}} B\) event, in which the transition-relevant subset is targeted rather than a global decomposition and the problem is treated dynamically. Within a selected transition corridor, barrier-crossing motion is extracted and microscopic players are ranked through a corridor-conditioned informativeness measure, after which the highest-ranked players are subjected to intervention tests of their transition leverage using the REST2 scaling scheme, metadynamics, and targeted mutation. Across three benchmark proteins, targeting these players facilitates basin-to-basin transitions while maintaining high fidelity to the prescribed transition corridor, providing handles for mechanistic analysis, enhanced sampling, control, and design.