Polarization of excitation light influences molecule counting in single-molecule localization microscopy

Single-molecule localization microscopy has been widely applied to count the number of biological molecules within a certain structure. The percentage of molecules that are detected significantly affects the interpretation of data. Among many factors that affect this percentage, the polarization state of the excitation light is often neglected or at least unstated in publications. We demonstrate by simulation and experiment that the number of molecules detected can be different from -40 up to 100 % when using circularly or linearly polarized excitation light. This is determined mainly by the number of photons emitted by single fluorescent molecule, namely the choice of fluorescence proteins, and the background noise in the system, namely the illumination scheme. This difference can be further exaggerated or mitigated by various fixation methods, magnification, and camera settings We conclude that the final choice between circularly or linearly polarized excitation light should be made experimentally, based on the signal to noise ratio of the system.