Skip to main content
Download PDF

Implementation of the Rank-Weighted Co-localization (RWC) algorithm in multiple image analysis platforms for quantitative analysis of microscopy images

Source Code for Biology and Medicine volume 11, Article number: 2 (2016) Cite this article

Abstract

Background

Quantitative co-localization studies strengthen the analysis of fluorescence microscopy-based assays and are essential for illustrating and understanding many cellular processes and interactions. In our earlier study, we presented a rank-based intensity weighting scheme for the quantification of co-localization between structures in fluorescence microscopy images. This method, which uses a combined pixel co-occurrence and intensity correlation approach, is superior to conventional algorithms and provides a more accurate quantification of co-localization.

Findings

In this brief report we provide the source code and implementation of the rank-weighted co-localization (RWC) algorithm in three (two open source and one proprietary) image analysis platforms. The RWC algorithm has been implemented as a plugin for ImageJ, a module for CellProfiler and an Acapella script for Columbus image analysis software tools.

Conclusions

We have provided with a web resource from which users can download plugins and modules implementing the RWC algorithm in various commonly used image analysis platforms. The implementations have been designed for easy incorporation into existing tools in a ‘ready-for-use’ format. The resources can be accessed through the following web link: http://simpsonlab.pbworks.com/w/page/48541482/Bioinformatic_Tools.

Introduction

The quantitative co-localization of markers in microscopy images has been widely used to study the spatial organization of intracellular environments. Traditional co-localization algorithms are based on either correlation of intensity values or pixel co-occurrence within regions of interest. We recently developed an algorithm that combines the information coming from intensity and pixel co-occurrence, and demonstrated in different scenarios that this rank-weighted co-localization (RWC) method produces superior results to traditional methods for quantification of co-localization [1]. While traditional co-localization methods that consider either pixel co-occurrence or intensity correlation have limitations, by integrating the two methods RWC provides a more reliable and accurate measure of describing the spatial distribution of two markers. Furthermore, the thresholding of images, which is commonly used to reduce noise in low intensity regions, often introduces bias as it is sensitive to the co-localization method chosen and how it is deployed. By contrast, RWC uses a weighting scheme to penalize low intensity regions in an image, thereby eliminating the need for manual thresholding and effectively reducing a major source of bias in quantification. We have shown that in a completely automated manner, RWC can be used to accurately quantify the spatial-temporal translocation of a cargo molecule as it passes through various organelles in the early secretory pathway. We have also demonstrated the use of RWC in improving clustering and classification of images [2, 3]. In this brief report, we present the implementation of the RWC algorithm in three different image analysis platforms, widely used by cell biology researchers. We believe that these implementations will be a valuable ‘easy-to-use’ resource for co-localization studies by the wider scientific community.

Implementations

JACoPx – an ImageJ plugin

ImageJ (http://imagej.nih.gov/ij/index.html) is an open-source, Java-based image analysis program developed at the National Institutes of Health [4]. It works with multiple operating systems (Windows, Mac OS, OS X, Unix-based systems) and its open architecture allows for extensions using custom Java plugins and macros. ImageJ is one of the most widely used image processing systems with applications in biological and medical sciences including analysis of microscopy, pathology and radiology images [5]. JACoP (Just Another Colocalization Plugin) is an ImageJ plugin that provides a variety of co-localization measures including Pearson’s coefficient, Overlap coefficient, Manders’coefficient and Costes’ automated threshold. We have extended JACoP to now include RWC coefficients. This extended plugin (JACoPx) provides an option along with the default measures to additionally calculate RWC coefficients. Since JACoP is already widely used, and provides an extensive collection of co-localization measures, we reasoned that implementing RWC into the same plugin will enable users to use the same familiar tool to also extract RWC coefficients, and thus also be able to easily compare these measures against other coefficients. JACoPx can be easily installed using the jacopx_.jar file provided through our website.

To install JACoPx:

  1. a)

    Download the associate jar file (jacopx_.jar) to the plugins folder within the ImageJ installation directory;

  2. b)

    Restart ImageJ. JACoPx should now be available for use under plugins tab.

MeasureCorrelationx – a CellProfiler module

CellProfiler (http://cellprofiler.org) is a Python-based, open source high-throughput image analysis system developed at the Broad Institute of MIT and Harvard [6]. CellProfiler is available for Mac OS X, Windows and Linux operating systems. CellProfiler has a modular design enabling users to choose the image processing routines specific to their assays. CellProfiler is a highly ranked cell image analysis tool that provides an interface to build analysis pipelines using the image processing modules as building blocks [7]. CellProfiler is designed for high-throughput analysis where quantitative phenotypic measurements can be extracted from thousands of images automatically. One of the measurement modules available in CellProfiler, called ‘MeasureCorrelation’, provides the Pearson’s co-localization coefficient for a pair of images. We have extended this module to include Overlap, Mander’s coefficient, Costes’ automated threshold and RWC coefficients. This extended module (MeasureCorrelationx) now enable users to extract a variety of co-localization measures with relative ease.

To install MeasureCorrelationx:

  1. a)

    Download the associated Python script (MeasureCorrelationx.py) to a folder;

  2. b)

    Point the “CellProfiler plugins directory” within the preferences option to this folder;

  3. c)

    Restart CellProfiler. MeasureCorrelationx should now be available under the measurement modules list.

RWC_Co-localization.script – an Acapella script for Columbus

Columbus Image Data Storage and Analysis System (http://www.perkinelmer.com/pages/020/cellularimaging/products/columbus.xhtml) is a proprietary web enabled system for storage and analysis of image data developed by Perkin Elmer. Columbus is a modular system similar to CellProfiler allowing users to build custom analysis pipelines and perform high-throughput analysis of very large image data sets. Modules are developed using Acapella scripts (Evotec Technologies GmbH), and Tony J. Collins (MacBioPhotonics, McMaster University, Canada) and the Andrews Lab (Sunnybrook Research Institute, Toronto, ON, Canada) have provided a suite of co-localization procedures (MBF_ColocalisationCoefficientsb03.proc) written in the Acapella scripting language. We have extended this procedure to create an Acapella script (RWC_Co-localization.script) that implements the RWC algorithm, such that this script can also be used as an independent ‘assay’ within the Columbus system. The script works with 3 channel input images and the user can select the two channels between which the co-localization coefficients wish to be calculated. The script allows for segmentation and detection of cell nuclei and cytoplasmic areas using the various inbuilt detection options, and then extracts co-localization coefficients on the objects identified. This script can be downloaded and directly imported as an assay in a ‘ready-to-use’ format.

Conclusions

We offer a web resource with implementation of the RWC algorithm in three of the most popular fluorescence microscopy image analysis software systems. While quantitative co-localization is a critical measure used in many conventional cell biology experiments, there is an increasing need for an analytical resource to perform such calculations on large-scale image data sets. We have provided open-source programs containing implementations of the RWC coefficient that can be easily incorporated into existing image analysis pipelines. All the software implementations are accessible via the following website: http://simpsonlab.pbworks.com/w/page/48541482/Bioinformatic_Tools.

References

  1. Singan VR, Jones TR, Curran KM, Simpson JC. Dual channel rank-based intensity weighting for quantitative co-localization of microscopy images. BMC Bioinformatics. 2011;12:407.

    Article  PubMed Central  PubMed  Google Scholar 

  2. Singan VR, Handzic K, Curran KM, Simpson JC. A method for improved clustering and classification of microscopy images using quantitative co-localization coefficients. BMC Res Notes. 2012;5:281.

    Article  PubMed Central  PubMed  Google Scholar 

  3. Singan VR, Handzic K, Simpson JC. Quantitative image analysis approaches for probing Rab GTPase localization and function in mammalian cells. Biochem Soc Trans. 2012;40(6):1389.

    Article  CAS  PubMed  Google Scholar 

  4. Schneider CA, Rasband WS, Eliceiri KW. NIH Image to ImageJ- 25 years of image analysis. Nat Methods. 2012;9:671–5.

    Article  CAS  PubMed  Google Scholar 

  5. Abràmoff MD, Magalhães PJ, Ram SJ. Image processing with ImageJ. Biophotonics Int. 2004;11(7):36–42.

    Google Scholar 

  6. Bolte S, Cordelières FP. A guided tour into subcellular colocalization analysis in light microscopy. J Microsc. 2006;244(3):213–32.

    Article  Google Scholar 

  7. Wiesmann V, Franz D, Held C, Münzenmayer C, Palmisano R, Wittenberg T. Review of free software tools for image analysis of fluorescence cell micrographs. J Microsc. 2015;257(1):39–53.

    Article  CAS  PubMed  Google Scholar 

Download references

Acknowledgements

We would like to thank the Carpenter Lab at the Broad Institute of Harvard and MIT, specifically Anne Carpenter, Thouis Jones and Lee Kamentsky for hosting and advising VRS when RWC was implemented in CellProfiler. We would also like to thank Tony J. Collins (MacBioPhotonics, McMaster University) for helpful discussions during the implementation of the RWC algorithm using Acapella scripts. This work has been partly funded by an Irish Research Council for Science, Engineering and Technology (IRCSET) graduate PhD scholarship in Bioinformatics and Systems Biology to VRS. The JCS lab is supported by a Principal Investigator (PI) award (09/IN.1/B2604) from Science Foundation Ireland (SFI).

Author information

Authors and Affiliations

  1. School of Biology and Environmental Science, Science Centre West, University College Dublin, Dublin 4, Ireland

    Vasanth R. Singan & Jeremy C. Simpson

  2. DOE Joint Genome Institute, Walnut Creek, CA, 94598, USA

    Vasanth R. Singan

Authors
  1. Vasanth R. Singan
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Jeremy C. Simpson
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to Jeremy C. Simpson.

Additional information

Competing interests

The authors declare that they have no competing interests.

Authors’ contributions

JCS conceptualized and designed the study. VRS designed the RWC algorithm and developed the software implementations. VRS and JCS wrote the manuscript. Both authors read and approved the final manuscript.

Authors’ information

VRS was a PhD candidate in the laboratory of JCS at University College Dublin, Ireland when this study was conducted. VRS is currently a Bioinformatics Systems Analyst at the DOE Joint Genome Institute in Walnut Creek, California, USA. JCS is the Professor of Cell Biology at University College Dublin, Ireland.

Rights and permissions

Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Singan, V.R., Simpson, J.C. Implementation of the Rank-Weighted Co-localization (RWC) algorithm in multiple image analysis platforms for quantitative analysis of microscopy images. Source Code Biol Med 11, 2 (2016). https://doi.org/10.1186/s13029-016-0048-8

Download citation

  • Received:

  • Accepted:

  • Published:

  • DOI: https://doi.org/10.1186/s13029-016-0048-8

Keywords

  • Fluorescence Microscopy Image
  • Image Processing Module
  • Cell Image Analysis
  • Image Analysis Pipeline
  • Image Analysis Platform

Source Code for Biology and Medicine

ISSN: 1751-0473