Bayesian conformational analysis of ring molecules through reversible jump MCMC

In this paper, we address the problem of classifying the conformations of m ‐membered rings using experimental observations obtained by crystal structure analysis. We formulate a model for the data generation mechanism that consists in a multidimensional mixture model. We perform inference for the proportions and the components in a Bayesian framework, implementing a Markov chain Monte Carlo (MCMC) reversible jump algorithm to obtain samples of the posterior distributions. The method is illustrated on a simulated data set and on real data corresponding to cyclo‐octane structures. Copyright © 2005 John Wiley & Sons, Ltd.