Need help or advice ? Come to the Icy club ! - Every other Thursday morning from 9h30 to 12h30 - Francois Jacob Building - Main hall - Pasteur

Associated publications

Don't forget the authors !
If you use this plugin, please cite it in your publication !

User reviews

5 star
4 star
3 star
2 star
1 star
average rating: 4.5

Please log-in to post a review
18 Jul 2016 22:49
Very nice to have the statistics missing from the colocalizer! I was wondering if for the object-based colocalization would be possible to also have the colocalized components, in order to display them on the original image?
30 Nov 2014 12:54
Very nice plugin regrouping various colocalization methods. I would love to use it in batch thanks to a protocol block.

Colocalization Studio

by Thibault Lagache

This plugin contains most of existing colocalization methods in 2D and 3D fluorescence microscopy images: Pearson and Manders coefficients, Image Cross Correlation Spectroscopy (ICCS) and Object-based methods.

Publication Id
See technical details
View complete changelog


If this plugin satisfies you, please cite: 

Lagache T, Sauvonnet N, Danglot L, Olivo-Marin JC. Statistical analysis of molecule colocalization in bioimaging. Cytometry A. 2015

This plugin contains most of the existing colocalization methods. Before using it, you can find here an introduction to colocalization methods.

This plugin is decomposed into two main method classes:

1- Correlation (pixel-based) methods.

You compute 3 correlation coefficients:

a- Pearson analysis. The p-value is obtained with pixel scrambling (we used the Central limit theorem instead of Monte-Carlo simulations)

Ref: Costes et al. Automatic and quantitative measurement of protein-protein colocalization in live cells. Biophys. J. (2004)

b- Manders analysis. The p-value is obtained with pixel scrambling (we used the Central limit theorem instead of Monte-Carlo simulations)

Ref: Manders et al. Measurement of co-localization of objects in dual-colour confocal images. J. Microscopy (1993)

c- ICCS analysis. Be careful: method very sensitive to noise and image shift. Image denoising (with appropriate filter and/or thresholding) is needed.

Ref: Comeau et al. A guide to accurate fluorescence microscopy colocalization measurements. Biophys. J. (2006)

2- Spatial (object-based) methods

This method relies on molecule detections with spot detector plugin. It then uses the Ripley's K function to analyze molecule spatial distribution. The statistical analysis of molecule colocalization is performed by the comparison of K function with critical quantiles (analytical formula as function of the ROI geometry and the number of spots). In particular, the maximum of the 0-mean and unit variance Ripley's K function is provided and can be directly compared with the quantiles of the Normal distribution. For example, if the max. of the K function is > 2,32 (which is the quantile at 99% of the Normal distribution (see the table), we can reject the null hypothesis of random distribution (meaning that molecules colocalize) with propbability 99%.

Ref: Lagache et al. A statistical analysis of spatial colocalization using Ripley’s K function Conf. Proc. IEEE ISBI (2013)

Then, the percentage and the distance between colocalized molecule is obtained with a parametrical fitting of the Ripley's K function. The option "Fit of the ripley's K function" provides the plot of the computed K function and the fit to check the robustness of the fitting procedure.

Ref: Lagache et al. Statistical analysis of molecule colocalization in bio-imaging. Cytometry A. (2015).