Using Nested EASE (nEASE)

Nested EASE (nEASE) is an extension of EASE. The nEASE algorithm includes a second, sub-level, iterative Fisher’s Exact Test on significantly enriched GO terms identified in a first-level EASE analysis. This sub-classification approach provides increased sensitivity for detecting enriched GO terms and thus affords a deeper understanding of possible mechanisms underlying a given condition under study. nEASE was added to MeV as a new feature for version 4.5.

Begin this tutorial after installing MeV.

Loading Example Data

Launch MeV by double-clicking on the TMEV.bat file (Windows), the tmev.sh file (Linux) or the MeV application icon (Mac OSX).

When you launch MeV, two windows will open. The small narrow window across the top of the screen is called the MeV main menubar. This window is used normally to open new MultipleArrayViewer windows and manage other MeV properties. We will not be using this menubar window for the purposes of this tutorial. The larger window that opens is called a MultipleArrayViewer (MAV). This is where the majority of our work will take place.

Download and unzip the file nease_example_files.zip. This file contains the expression data and supporting GO term files we will use to replicate the analysis presented in the manuscript.

Choose File-> Open Analysis from the MAV window. In the file chooser that opens, navigate to the folder where you unzipped nease_example_files.zip, and select the file ER_status_SAM_1_Miller.anl and choose Open. A saved analysis will be loaded into MeV. This may take some time.

 

These data are fully described in the manuscript. They are based on the data from Miller et al. (2005) and Minn et al. (2005). 

Choosing Nested EASE Parameters

EASE is found under the Meta Analysis menu itemEASE is found under the Meta Analysis menu itemNear the top of the MAV is a row of colorful drop-down menus. These menus contain the analysis options available in MeV.

After loading the analysis file, click the Meta-Analysis drop-down menu, and select EASE Cluster Analysis. An initialization dialog will appear.

 

 

 

 

Step 1: Selecting the EASE file system

To manually load an EASE file system click the button marked Custom in the EASE Annotation Analysis dialog which will bring up the EASE Advanced Parameters dialog, then click the Browse button. Proceed to navigate to the folder you previously downloaded, nease_example_files. Inside it, open the folder named data\ease\affy_HT_HG-U133A_EASE. You should see three folders inside, Data, Enhance and Lists. Do not select any of these folders. Simply click Open.

 

Step 2: Population Selection EASE Advanced Parameter DialogEASE Advanced Parameter Dialog

Make sure the option Select Background Population from File is selected in the EASE Advanced Parameter dialog. Click the button labeled File Browser in the Population panel and select the file Lists/affy_HT_HG-U133A/Populations/ProbesetIDs.txt.

 

 

 

 

 

 

 

 

 

 

 

 

Step 3: Annotation Parameter Selection

From the drop-down menu labeled Annotation Key in the MeV Annotation Key panel, select the heading PROBE_ID. Click the checkbox labeled Use Annotation Converter. Click the button labeled File Browser. The file selection dialog that opens should already be set to the correct directory. (If it is not, navigate to nease_example_files\affy_HT_HG-U133A_EASE\Data\Convert.) Select the file affy_HT_HG-U133A_ProbesetIDs.txt and click Open. Under the heading Gene Annotation /Gene Ontology Linking Files, click the Add Files button. Again, the resulting file browser should already be displaying a list of the available GO term files. (If not, navigate to nease_example_files\affy_HT_HG-U133A_EASE \Data\Class directory within the MeV folder.) Hold down the control key and click on the files GO Biological Process.txt, GO Molecular Function.txt and GO Cellular Component.txt to select them. When the three files are selected, click Open.

 

Click the OK button to return the the main EASE dialog.

 

Step 4: Statistical Parameter SelectionnEASE is an intensive module and will take some time to runnEASE is an intensive module and will take some time to run

Click to select the check box next to Run Nested EASE, near the bottom of the EASE Annotation Analysis dialog. Now click the Ok button to run EASE and Nested EASE. This will take some time. 

Viewing the nEASE results

The standard EASE analysis, as described in Hosack, et al. (2003), will run, followed by the Nested EASE analysis. The nEASE results are included as part of the standard EASE results, as a subnode on the result tree, labeled Nested EASE. Double-click the result node labeled EASE Analysis to see the nested ease results appear. We recommend that you expand the size of the window containing the Result Tree by clicking and dragging on the dividing bar between the Result Tree and the larger viewing window on the right. Click on the result nodes to explore the result data within.

 

Nested EASE Summary Table.Nested EASE Summary Table.

The most useful view is found in a node labeled Nested EASE Summary Table, which contains the summarized data for all of the nEASE results. These are the data reported in the manuscript. To examine the genes which drive a result row in the nEASE Summary Table, left-click to select a row, and right-click to open up a context-sensitive menu. Choose Open Viewer -> Expression Image. MeV will open a heatmap view of the probe expression values that correspond to the genes that were members of that group. 

 

 

View the expression profiles for the genes in the EASE group of interest.View the expression profiles for the genes in the EASE group of interest.

All of the usual heatmap-related options are available in this window, including adjusting the color display, changing the gene annotation displayed, and storing the displayed genes in a new cluster. Note that the left-hand panel, the Result Tree, does not obviously change when this heatmap is opened. However, scrolling the Result Tree down will reveal that a new result node has been selected, corresponding to the Nested EASE result that was selected from the Summary Table. 

 

References D. A. Hosack, G. Dennis, B. T. Sherman, H. C. Lane, R. A. Lempicki, Genome Biol 4 (2003). A. J. Minn, et al. , Nature 436, 518. (2005) L. D. Miller, et al. , Proceedings of the National Academy of Sciences 102, 13550 (2005).