Colocalization in Volocity

When using two or more fluorescent channels you may have areas where the dyes overlap. The Colocalization features in Volocity can be used to evaluate this. Colocalization in Volocity calculates measurement statistics based on the voxels and their intensities, it is designed specifically for Colocalization analysis as described by Manders et al (1993).

Let's look at Colocalization in Volocity with an example dataset. By the way, this information is all available in the Volocity User's Guide which you can download here.

Before we start we will need a two channel dataset, both channels will need to have the same image depth. Let's take a look with this example image sequence.

Click on the "Colocalization" tab to switch to the "Colocalization" view.

You will need to select the channel X and channel Y using the pop-ups at the top of the dialog.

The image on the left side of the dialog is an extended focus merge of the dataset.

Below this Volocity has calculated measurement statistics based on the intensity information in the whole image.

As you can see there is a scatter plot is on the right side of the dialog. Each axis represents the gray levels in each of the channels of the image sequence we are analysing.

Just below the scatter plot are the thresholds these allow you to set both upper and lower thresholds for each of the channels.

We need to work with the scatter plot to set the lower thresholds so that they are above the background intensities for each channel.

There are a number of techniques you can use to set the thresholds.

1.  Draw a region of interest in a background area of the preview window. Go to the "Colocalization" menu … "Set Thresholds from ROI".  Volocity calculates the mean intensity of the region for each channel and uses this as the lower threshold.

This is just to give you an idea, the region of interest I have drawn is not exactly in the background so I will need to use one of the other techniques.

2. The threshold values can be entered manually or you can use the slider controls on the side of the scatter plot. See below how the scatter plot updates to show that the voxels have been excluded.

 With this dataset I know what background intensities I would like to use so I will enter them directly into the minimum threshold boxes for each channel.

 

3. From the "Colocalization" menu "Automatic Thresholding" can be selected. Volocity uses statistical tests to automatically generate thresholds as described by Costes et al. (2004)

When "Automatic Thresholding" is selected the manual setting of thresholds and "Set thresholds from ROI" are disabled.

If you have a multiple timepoint dataset Volocity will recalculate the thresholds for each timepoint that is being viewed. To change the current timepoint use the Navigation palette.

Any voxels that fall outside the threshold ranges will be removed from both the scatter plot and the image preview for that channel.

If you fail to set the thresholds, voxels where both images show background areas may be interpreted as being colocalized.

Next we need to select the voxels to be analysed on the scatter plot.

This is done by drawing a region(s) of interest on the scatter plot.

You can select the entire scatter plot by going to the "Edit" menu … "Select All". "Select none" can be used in this menu to do just that and remove any regions on the scatter plot.

If you need to you can draw a sub region on the scatter plot for analysis, use the ROI tools to do this.

As regions are drawn the preview will update to display the location of the selected colocalized voxels. The statistics will also update based on the selected voxels.

If you don"t like the look of the colocalization overlay in the image view go to the "Colocalization" menu … "View Options".

Here you can change the color and opacity of the feedback.

That looks a bit better.

Next we need to look at the Colocalization output, there are two methods of doing this.

You can use Volocity to "Generate Colocalization".

Select "Generate Colocalization" from the "Colocalization" menu.

The Field statistics measurements are measurements representing the whole volume rather that the objects within it.

As we only have one timepoint we will get only one measurement containing all of the colocalization statistics.

If the option is selected the scatter plot will be saved to a folder in the library, the scatter plot is a 2D image and can be easily exported.

If selected a colocalization channel can be made from the colocalized voxels. The channel is produced via a brightest point merge making use of the color and intensity of the voxels in the original channels.

As an alternative to using "Generate Colocalization" you can use "Measure Objects" from the "Tools" menu. This will allow you to look at the colocalization coefficients Mx and My for each object.

You can use classifiers to perform this but let's first approach this by obtaining colocalization measurements from a specific region in the preview.

Draw ROIs in the Preview of the Colocalization window.

As you can see I have draw six regions

Go to the "Tools" menu … "Measure Objects"

As in the dialog above select "Colocalization" as the measurement type. We are also measuring the items from the current selection. When ready click on the "Measure" button.

The measurement table will contain a measurement for each region that was drawn in the preview.

 

See we have six measurements in total.

The Classifier can be used to select the objects of interest. A classifier can be made whilst you are viewing the "Colocalization" view or in the "Image" view if this is what you are more familiar with.  Please note however that the in order to obtain the Colocalization measurement type in the "Measure Objects" dialog,  the "Colocalization" view must be active.

In the "Measure Objects" dialog select the classifier you would like to use, target the channel and remember to select the measurements that you require.

If the "Colocalization" measurement type is selected the colocalization coefficients (Mx and My) will be calculated for each object identified by the classifier.

There are some important considerations to take into account when using colocalization, for further details please consult the Volocity User's Guide.

I hope this has provided an idea of what can be achieved with colocalization, see you next month.

References
M.M. Manders, P. J. Verbeek & J. A. Aten
Measurement of co-localization of objects in dual color confocal images
Journal of Microscopy. Vol. 169, Pt 3, March 1993, pp. 375-382
S. V. Costes, D.Daelemans, E. H. Cho, Z. Dobbin, G. Pavlakis & S. Lockett
Automatic and Quantitative Measurement of Protein-Protein Colocalization in Live Cells
Biophysical Journal Vol. 86, 2004, pp. 3339-4003