SMITE (single molecule imaging toolbox extraordinaire), a MATLAB package.

Abstract

We present SMITE, a MATLAB package that features analysis tools for fluorescence single molecule imaging with an emphasis on single molecule localization microscopy (SMLM) and single-particle tracking (SPT). Some C and CUDA code is embedded in the MATLAB infrastructure to provide CPU/GPU speed-ups for particularly expensive computations. SMITE is organized around two data structures: SMF (Single Molecule Fitting), containing parameters that completely define the data analysis, and SMD (Single Molecule Data), which contains the results of the analysis. The code is divided into namespaces and subclasses that group related concepts and functionality together. SMITE is designed so that low-level tools as well as higher-level application specific interfaces are easy to use, making use of the SMF and SMD data structures whenever possible. Interaction with these tools is via GUIs or scripting. Extensive internal documentation of the interfaces and many usage examples are provided to aid the user. Code coverage includes mature SMLM data analysis techniques (applying gain and offset corrections to raw data, finding localizations, thresholding localizations based on various criteria, frame connection and drift correction), SMLM/SPT simulations, sophisticated SPT analyses, post-processing clustering and statistical analyses (e.g., Bayesian grouping of localizations for removing repeat localizations from the same emitter, diffusion analysis, hidden Markov models for characterizing dimers in SPT results), a variety of visualizations, experimental point spread function creation and characterization. SMITE is available via GitHub: https://github.com/LidkeLab/smite