Scientists from Arizona State University are leveraging the tools of data science to study molecular activity more quickly than is possible through traditional fluorescence correlation spectroscopy (FCS).
While FCS provides estimates of dynamical quantities, it requires high signal-to-noise ratios and time traces that are typically in the minute range. The researchers at ASU are using Bayesian analysis to overcome the limitations of fluorescent correlative methods in using short and noisy time traces to deduce molecular properties such as diffusion coefficients.
Using Bayesian nonparametrics for the direct analysis of the observed photon counts emitted by single molecules, the researchers were able to analyze time traces that are too short to be analyzed by existing methods, including FCS. The team’s new analysis approach could extend the capability of single molecule fluorescence confocal microscopy techniques to probe molecular processes at a speed that is several orders of magnitude faster.