Overview Code and example data associated with 'High-density volumetric super-resolution microscopy' by Daly et al. including: 3D reconstruction code for single molecule light field microscopy, 3D single-particle tracking code, Code to compare single molecule data to ground truth coordinates, and Example raw SMLFM datasets. Single molecule light field microscopy reconstruction Authors: Ruth R. Sims, Kevin O'Holleran, Sohaib Abdul Rehman, Ezra Bruggeman and Sam Daly Objective: It takes in 2D localisation data (x,y) captured on a fourier light field microscope and turns it into 3D localisations (x,y,z). It does this by first assigning (x,y) to (x,y,u,v) space and using microscope parameters to calculate z position via parallax. Example data has been included (example_data_bcr_fixed_bcell.csv), which comprises 5,000 frames of a whole B cell membrane captured using single molecule light field microscopy. For more information, see: https://doi.org/10.1364/OPTICA.397172 Single-particle tracking code Groups 3D coordinates from a .csv file (i.e. [x y z frame]) into molecular trajectories based on user defined metrics. This code incorporates track.m by John C. Crocker (University of Chicago) and MLEs by Joseph S. Beckwith (University of Cambridge). For further details see the description within the tracking functions. Example 3D data can be found in the folder 'example_data_bcr_tracking' that comprises some single particle tracking data collected from live B cells. The parameters contained in the distributed code enable instant 3D reconstruction of single trajectories of BCRs. Matching codes These scripts match fitted 2D/3D localisation datasets derived from simulations with their ground truth coodinates to determine positive predictive value and sensitivity for different 3D PSFs at increasing density. Example raw localization datasets have been included for the standard PSF. To begin, run Standard.m and select the directory containing subfolders 'numEmittersPerFrame_X'. Example Raw Data Example raw SMLFM datasets to accompany the experimental component of the article. All three localization datasets (200 frames) have a pixel size of 266 nm, a bias of 400 counts, and a conversion gain of 40 counts/photon.

Supported by The Royal Society (RGF\EA\181021)