STAC stands for "Significance Testing for Aberrant Copy-Number". STAC tests for significantly concordant aberration across multiple samples in array CGH data.
This document walks through a couple of examples. This assumes you have STAC running, for installation instructions see the STAC user's guide.
After starting up STAC Station, it should look like this:
We are first going to open a file in location format on which to perform STAC analysis: sampleinput1.txt. Right-click the link and save the file somewhere on your local disk. (Mac users can't right-click, but save the file to your disk using whatever Mac commands are necessary).
For best results STAC analysis should be performed on a single chromosome arm at a time
(click here for more notes on formatting and pre-processing).
To open the file to perform STAC analysis, click on
Find the file sampleinput1.txt that you've saved on your local disk. Notice that once you've opened it, the data and filename are displayed and the "Analyze" button on the task bar becomes enabled.
Hit the "ANALYZE" button to start the analysis - the progress is displayed in the progress bar at the bottom of the window.
Once the analysis is completed the results are automatically saved to the disk as a file with the same name as the input file but with a ".stac" suffix added. This is the file being viewed as indicated in the info bar "Currently viewing: ...".
There are two statistics used to assess significance. We'll first look at the frequency confidence. There are two frequency confidence buttons. Click on the first one on the left.
The difference between the two buttons is that the first one displays the confidence with a bar graph and the second displays it with a line graph. STAC Station displays "confidences" instead of p-values, the confidence is simply one minus the p-value. The higher the bar, the higher the confidence at that location. The confidence scale is given along the right edge of the graph.
The footprint confidence buttons are just to the right of the frequency confidence buttons. Hit the footprint confidence bars button and it will display the footprint confidences combined with the frequency confidences which are already turned on. Notice how the footprint confidence picks up a couple of tightly aligned stacks of low frequency that the frequency confidence misses.
Notice that some bars are darker than others. These are the bars that exceed confidence of .95. You can change the .95 cutoff using the Options Display Enter Confidence Cutoff
Change the confidence cutoff to .99 and notice how fewer bars are now shaded darker:
Let's zoom in a bit so we can see better. There are five buttons used to change the size of the display. You can stretch things both vertically (using the "Exp IDs" buttons), or horizontally (using the "Loc IDs" buttons), or you can fit the display to the screen (using the "fit to screen" button).
Increase the location ID font size by hitting the "Loc IDs larger" button five or six times:
The result should look like this:
Now let's highlight the intervals that are involved in the confidence at position (5). Choose the
Enter "5" in the text box. This will highlight in blue all the relevant intervals for position 5:
In this case every interval covering position 5 is highlighted, but this might not be the case, generally it is only a subset of the intervals which contribute to the confidence at a given position.
Now let's export the results to a tab delimited text file.
Save the report file as sampleinput1.STAC.report. This is a tab delimited file that gives the frequency and footprint confidences for each location.
Note that you can also save the image as a jpeg.
Now, the STAC result file has automatically been asved so it can be reopened later for re-viewing, without having to rerun the analysis. You can later reopen using the menu item File Open to View File.