Getting Started

All you really need to get started is some FCS data. There are only a few user-selectable parameters that govern QTube's behavior (see “Setting Parameters”), and the default parameters are appropriate for many applications. So we encourage you to just try it out!

Remember that QTube wants to compare fingerprints for parameters that are common to two or more tubes, where the data for each tube is stored in its own FCS file.



At the top of QTube there are the familiar MenuBar and ToolBar.  Hovering over a button in the ToolBar will display the function of the button.  At the very bottom is a giant button that will compose an email to Cira from you to discuss licensing for use outside of your organization, if you wish.  In between, the screen is divided into two sections by an adjustable divider.  In the upper one will be a list of parameters that QTube finds in your FCS files.  You will use the table that appears here to select parameters you want included in the fingerprints.  The results will appear in the lower one.  

So, after launching the QTube application, click on “File -> Select FCS Files...” or on the left-most button in the Toolbar [ ].



A File selection dialog will appear. Browse to the directory that contains the files you want to analyze (QTube will remember this location for subsequent runs, making it convenient to come back to the same or nearby location later).

You can select multiple files in several ways. If the files are listed contiguously you can click on the first file of a group, and then shift-click on the last one to select the entire group. Or, you can CTRL-click on each file to add it to the list. Regardless of which method you choose, once all of the files that you want analyzed have been selected, click on the “Open” button. This returns you to the QTube main panel.

Note that the “Run” button [  ] is disabled. That's because Qtube doesn't yet know which parameters you wish to be compared. QTube at this point has analyzed the FCS headers and determined which parameters are common across all of the tubes based on the FCS Standard $PnN and $PnS parameter names assigned by your cytometer. These parameters are shown in black text – the parameters that differ across the tubes are shown, but are disabled to avoid inadvertent use of parameters that should not be compared because different reagents are used for the same parameter in different tubes.

So, select the parameters you wish compared by clicking in the appropriate check-boxes.




In the example above, we have selected two of the three available parameters: 2 (Side Scatter) and 5 (CD45-PerCP). 

Now just click the Run [ ] button. In a few seconds (depending on the speed of your computer) the results will be displayed in the panel below.



QTube presents the results for several “metrics” in a color-coded table.

Note on Memory

If the total number of events in the set of files you've chosen is large, QTube will sample at random an equal number of events from each file.  This is to enable QTube to run within a small memory footprint on your computer.  If this happens, QTube will inform you, and tell you how many events from each file it's using.

Metrics

We've included two different metrics for measuring similarity, each with somewhat different statistical properties. You can be quite sure that if both metrics for all tubes are “in the green” that your data are highly self-consistent (that is, unless you've gotten threshold parameters badly boluxed up!).

Each metric is computed from the same fingerprint representation. The fingerprint is expressed as an array of values, each of which corresponds to a subregion in the space you specified by your choice of fingerprinting parameters. Each value is the ratio of the observed probability in the subregion divided by the expected (uniform) probability.

max - The max metric is computed by first taking the log2 of the relative densities. Thus, if there are twice as many events than expected in a subregion, the value will be +1.0. If there are half as many as the expected number it will be -1.0. The max metric is then the maximum absolute value of these transformed densities. This best represents the extreme deviation from the norm. It is probably the most sensitive of the metrics in that it is sensitive to a localized abnormality.

Standard Deviation - This metric is simply the standard deviation of the relative densities in the fingerprint. It provides information regarding the distribution of values around the expected value. Like the max metric, values close to zero mean that the data are very consistent. Unlike the max metric, a value of 1 means something a little more complicated, and a precise interpretation would require examining how the values are distributed. In general these values should be considerably smaller than the max metric values.

Resolution

The resolution parameter governs the scale to which Cytometric Fingerprints are taken. Essentially, we sub-divide events into relatively smaller subregions of multi-parameter space at the higher resolutions. If you only have two or three common parameters, low or medium resolution should be adequate. If you are comparing multiple tubes on more than three parameters (an unusual circumstance) you may want to consider using High resolution. Higher resolution will take a few seconds longer to run, depending on the speed of your computer and the size of your FCS files.