Spatiotemporal Clustering of Repeated Super-Resolution Localizations via Linear Assignment Problem

Area of Science:

• Biophysics
• Optical Microscopy
• Computational Biology

Background:

• Super-resolution microscopy techniques like dSTORM, PALM, and DNA-PAINT generate datasets with multiple localizations from single emitter blinking events.
• These repeated localizations complicate data analysis and underutilize fluorescence photons.
• Current methods for combining localizations use fixed thresholds or Gaussian error assumptions, which can be suboptimal.

Purpose of the Study:

• To develop an improved clustering method for super-resolution microscopy data.
• To address the challenge of handling multiple localizations from single blinking events.
• To enhance the accuracy and efficiency of super-resolution data analysis.

Main Methods:

• A novel clustering approach is presented that integrates localization precision, local emitter density, and a kinetic blinking model.
• This method optimizes the connection of spatiotemporally colocated localizations originating from the same emitter.
• The approach moves beyond traditional distance/time thresholds and Gaussian error assumptions.

Main Results:

• The developed method effectively clusters repeated localizations arising from single blinking events.
• It provides a more accurate representation of emitter positions by accounting for localization precision and blinking kinetics.
• This leads to a more complete utilization of fluorescence photons within the dataset.

Conclusions:

• The proposed clustering method offers a significant advancement in analyzing super-resolution microscopy data.
• It enhances the interpretation and analysis of datasets from techniques like dSTORM, PALM, and DNA-PAINT.
• This approach improves the efficiency of photon usage and the accuracy of single-molecule localization.