Bayesian-inference-based fluorescence correlation spectroscopy and single-molecule burst analysis reveal the...

Fluorescence correlation spectroscopy (FCS) is a powerful tool to gain information about dynamics of biomolecules. However, the key problem is to extract the rates hidden in the FCS data by fitting the data to a meaningful model. A number of different fitting approaches have been described in recent years but the extraction of relevant information to date has still been limited by numerous experimental problems and the fact that the set of starting parameter values chosen could often predefine the result. We establish a new way to globally analyze FCS data based on Bayesian inference to overcome these issues. Moreover, the influence of other remaining experimental error sources, for example, photophysics, is excluded by additional means. Using this approach in combination with the results from single-molecule burst analysis, we investigate the kinetics of DNA hairpins labeled with a variety of different fluorescent probes as a function of the salt concentration. We find that the rates of hairpin opening and closing as well as the equilibrium constant of the transition depend on the characteristics of the dye molecules used to label the hairpin. Thus, great caution has to be used when utilizing dye molecules as reporters for the kinetics of dynamic macromolecular structures.