Deconvolution of High-Dimensional Ion Mobility Data Using Reversible-Jump Markov Chain Monte Carlo

Deconvolution of multicomponent arrival time distributions is known to be a highly challenging task due to the "curse of dimensionality". Therefore, the development of a robust global optimizer is a crucial milestone in the automated analysis of arrival time distributions for collision cross section extraction and population analysis. Here, we report an approach that combines gas-phase ion transport theory with equi-energy sampling, Bayesian sequential partitioning, and reversible-jump Markov chain Monte Carlo to automatically determine probabilities of deconvolution solutions and predict the number of components in the arrival time distribution. The robustness of the method was evaluated against synthetic and experimental drift tube ion mobility data to find global deconvolution solutions and automatically determine collision cross sections. Analyzing the collision-induced unfolding profile of cytochrome c using the developed pipeline ultimately enables tracking of the conformational ensemble under activation. Additionally, using the same strategy on isomeric O-glycan species further revealed the potential for cross-platform collision cross section annotations.