The toroidal nucleus is a novel chromosomal instability (CIN) biomarker which complements the micronucleus. Understanding the specific biological stresses leading to the formation of each CIN-associated phenotype requires the evaluation of large panels of biological images collected from different genetic backgrounds and experimental conditions. However, the quantification of toroidal nuclei is currently a manual process which is unviable on a large scale. Here, we present QATS (QuAntification of Toroidal nuclei in biological imageS), a tool that automates the identification of toroidal nuclei, minimizing false positives while highly agreeing with the manual quantifications. Additionally, QATS identifies micronuclei for a convenient comparison of both CIN biomarkers. QATS is an open-source ImageJ plugin with a user-friendly interface that enables a wide scientific community to easily assess the frequency of CIN biomarkers for the determination of CIN levels in cellular models.
We provide the following files (click to download):
QATS.tar.bz2 which includes (file is <4Mb):
QATS_plugin.java: the Java source code.
QATS_plugin.class: the compiled Java code.
input/image_test.tif: test image to use as input for QATS.
output/: folder containing the expected output after running QATS with the provided example.
QATS_user_manual.pdf: this user manual.
QATS_examples.tar.bz2 which includes (file is ~180Mb):
input_images/: folder containing a diverse set of 100 images to test QATS.
output/: folder containing the expected output.
Installation
QATS is an ImageJ plugin implemented in Java. It has been tested on ImageJ 1.54f in both Windows 10 and Ubuntu 22.04. ImageJ can be downloaded from https://imagej.net/ij/index.html. To update ImageJ, open the program, click on “Help → Update ImageJ”.
To install QATS, copy the file QATS_plugin.class to the ImageJ/plugins folder and open ImageJ, or open ImageJ and click on “Plugins → Install” and select the file QATS_plugin.java. In both cases, QATS_plugin should be shown under “Plugins”.
Running
To run QATS, open ImageJ and click on “Plugins → QATS plugin”. Next, select the parameters and running mode. QATS accepts both jpg and tif files as input.
Parameters
We recommend using the automatically calculated parameters and to not modify the circularity values. However, users can also indicate the size ranges for the particles by clicking on “Manual parameters”.
There are 12 parameters describing the size (in square pixels) and circularity of nuclei, toroidal nuclei, and micronuclei that the plugin will identify. Following, there is a description of these parameters:
Nucleus size (min): minimum size of the nucleus (NUC_SIZE_MIN)
Nucleus size (max): maximum size of the nucleus (NUC_SIZE_MAX)
Nucleus circularity (min): minimum circularity of the nucleus (NUC_CIRCULARITY_MIN)
Nucleus circularity(max): maximum circularity of the nucleus (NUC_CIRCULARITY_MAX)
TN size (min): minimum size of the toroidal nucleus (TN_SIZE_MIN)
TN size (max): maximum size of the toroidal nucleus (TN_SIZE_MAX)
TN circularity (min): minimum circularity of the toroidal nucleus (TN_CIRCULARITY_MIN)
TN circularity (max): maximum circularity of the toroidal nucleus (TN_CIRCULARITY_MAX)
MN size (min): minimum size of the micronucleus (MN_SIZE_MIN)
MN size (max): maximum size of the micronucleus (MN_SIZE_MAX)
MN circularity (min): minimum circularity of the micronucleus (MN_CIRCULARITY_MIN)
MN circularity (max): maximum circularity of the micronucleus (MN_CIRCULARITY_MAX)
The plugin will only identify as nuclei those particles of size between NUC_SIZE_MIN and NUC_SIZE_MAX, and with a circularity between NUC_CIRCULARITY_MIN and NUC_CIRCULARITY_MAX. Similarly, the plugin will only identify as toroidal nuclei those particles of size between TN_SIZE_MIN and TN_SIZE_MAX, and with a circularity between TN_CIRCULARITY_MIN and TN_CIRCULARITY_MAX. Importantly, these particles should be inside the identified nuclei and with an intensity similar to the background. Finally, the plugin will only identify as micronuclei those particles of size between MN_SIZE_MIN and MN_SIZE_MAX, and with a circularity between MN_CIRCULARITY_MIN and MN_CIRCULARITY_MAX, which are relatively close to an identified nucleus. It should be noted that each nucleus can only be associated with a single micronucleus.
The option “Highlight all nuclei” will result in an output image in which the processed nuclei will be highlighted. The option “Report nuclei ID” will generate an output image in which the processed nuclei will have an ID that will map to a spreadsheet containing topological features.
Next, select whether to run QATS on a single image file or on all images present in a folder. QATS accepts both jpg and tif files as input. QATS can also run on an already open image.
Output
The results of the plugin will be located in a subfolder inside the input folder. The plugin will store the results in the subfolder “QATS_output.1”. If that subfolder already exists, it will add a unit to the subfolder suffix (QATS_output.2, QATS_output.3, …) until it finds a subfolder that does not exist. For each image quantified (for instance, XXXX.tif), QATS will generate:
An output image highlighting the identified toroidal nuclei (TN, in magenta) and micronuclei (MN, in green). The header of this image will contain the file name and the number of identified TN and MN. The filename of this image will be “quantification_XXXX.tif”. If “Highlight all nuclei” was selected, the identified nuclei will also be highlighted. If “Report nuclei ID” was selected, a unique number will be shown on top of each identified nuclei. These IDs map to the coordinates spreadsheet (see below).
A spreadsheet with the total number of nuclei (both considering and disregarding those nuclei at the edges), micronuclei, and toroidal nuclei. The filename will be “quantification_XXXX.csv”.
A spreadsheet containing the coordinates of all the identified nuclei, including size and shape descriptors, and whether they have a toroidal nucleus and/or micronucleus. The filename will be “coordinates_XXXX.csv”.
In the case of running the plugin on a folder with several images, the plugin will generate a spreadsheet summarizing the results with a line for each image and the number of nuclei, micronuclei, and toroidal nuclei identified. These results will be in a file called “output_summary.csv”.
It should be noted that using different versions of ImageJ may result in slightly different results. Total running time for the 100 images on a laptop with a i7-1165G7 processor was ~90 seconds.
For further information you can contact us at carles.pons@irbbarcelona.org and caroline.mauvezin@ub.edu
