engleski

Simple model of quantitative co-localization: from tissues to single molecules

Introduction: Fluorescence microscopy is a widely used tool in biological and material studies, providing imaging contrast that outperforms conventional optical imaging techniques. Additionally, with the use of fluorescence dyes, multicolor imaging of various cell components or even single molecules is possible. The analysis of such images is most often done qualitatively, which significantly limits the comparability of results obtained for different samples. In this work, we propose a simple and straightforward method for quantitative analysis of co- localization of up to three image channels. Methods: The algorithm was implemented in Python3. Python is developed under an OSIapproved open- source license, making it freely usable and distributable. Source files as well as binary packages are obtainable on the Bitbucket repository, and are available for tests and further development. Results: We prepared an application that allows massive data processing of hundreds of images, corresponding to different cases. This requires proper filename convention, which is crucial for the proper identification of image channels. The algorithm requires only one parameter - a threshold value, to determine the background level. Logical AND operation is applied for every co-localized pair, the result is quantified and referred to the input data. Venn diagrams are plotted for all channels as well as thresholded and output images. Finally, the report file is generated in a spreadsheet document format. Discussion: To compare a large number of obtained results, a 2D vector plot is generated to graphically illustrate the statistical relevance of co-localized channels. In this manner, the results obtained for tens of samples composed of three different channels each could be plotted on a single picture. Conclusion: Despite the simplicity of the proposed method, we believe the proposed algorithm and graphical representation of results can simplify the process of data analysis and provide a fast and versatile procedure for quantitative image co-localization.

co-localization, single molecules, imaging, algorithm

nije evidentirano