Evaluation of Fluorophore Counting Approaches in Superresolution Microscopy

Abstract

The recent invention of superresolution fluorescence microscopy has revolutionized the way we study biological structures at the cellular to molecular levels. However, extracting quantitative information, such as stoichiometry, from these new tools remains challenging, due to the stochastic photo-physical properties of fluorescent labels. In this work, we characterize the accuracy of two approaches, time-delay and maximum likelihood estimation, in determining the number of fluorescent emitters in simulated time traces. We present regimes where the time-delay method estimates the fluorophore count accurately, and inaccurately. We show that the maximum likelihood approach gives more accurate estimates of number of fluorescent emitters, and we characterize the sensitivity of this approach to perturbations in the transition rates. These efforts serve as a starting point to further improve quantitative analysis at superresolution data and provide guidance for experimental verification.