A Computational Method for Tracking and Analysing Dynamic Phenotypes of Individual Macromolecules

Fluorescence time-lapse microscopic imaging is a powerful tool for investigating dynamic cellular processes. To date, automated fluorescence microscopes are increasingly used to generate high-content spatial-temporal image data at high-throughput. Such large-scale data require much computational effort to quantify complex macromolecular dynamics such as temporal changes in fluorescence intensity, spatial locations, trajectory and velocity. Here, we propose a semi-automated method for tracking dynamics of individual macromolecules and performed quantitative analysis of the extracted phenotypes. We compared our proposed method with the manual identification of macromolecules and with a particle tracking method provided by the image analysis software, ImageJ. Our results clearly provide more object trajectories than the ImageJ method and our phenotypic analysis illustrates the relationship between two significant features with human-interpretable visualization.