Size: 6258
Comment:
|
Size: 6258
Comment:
|
Deletions are marked like this. | Additions are marked like this. |
Line 16: | Line 16: |
* EX_NT.cls : phenotype definition for expression file required by GSEA. | * ES_NT.cls : phenotype definition for expression file required by GSEA. |
Enrichment Map GSEA Tutorial
Contents
Outline
This quick tutorial will guide you through the generation of an Enrichment Map for an analysis performed using GSEA Gene Set Enrichment Analysis,
To run this tutorial:
- You need to have Cytoscape installed : minimally 2.6.3 must be installed but preferable to have the latest version of cytoscape
- Install the Enrichment Map plugin from the Cytoscape plugin manager. If you install it manually (e.g. if you need to install a new version that doesn't happen to be in the plugin manager yet), then it must be in the Cytoscape-[Version#]/plugins folder
You need to download the test data: GSEATutorial.zip
Description of the tutorial files contained in the !GSEATutorial folder:
- ES_NT.cls : phenotype definition for expression file required by GSEA.
- MCF_ExpMX_v2_names.gct : Expression File - Estrogen treatment, Official Gene Name as key. - Data for 12hr,24hr and 48hr.
Human_GO_AllPathways_no_GO_iea_April_15_2013_symbol.gmt: Gene set definition file.
Instructions
Step 1: Generate GSEA output files
GO to GSEA website - http://www.broadinstitute.org/gsea/
- Click on Downloads in the page header.
From the javaGSEA Desktop Application right click on Launch with 1 Gb memory.
- Click on “Save Taget as…” and save shortcut to your desktop or your folder of choice so you can launch GSEA for your analysis without having to navigate to it through your web browser.
- Double click on GSEA icon you created.
Click on Load data in left panel.
Click on Browse for files… in newly opened Load data pane.
Navigate to directory where you stored tutorial test set files. Select raw expression (.gct) file, sample class file(.cls) and gene set (.gmt) file. Click on Open.
Wait until confirmation box appears indicating that all files loaded successfully. Click on Ok.
Click on Run GSEA in left panel.
Click on the arrow next to the Expression dataset text box. Select the expression set you wish to run the analysis on.
Click on … next to the text box of Gene Set Database.
Click on Gene Matrix (local gmx/gmt) tab.
Select gmt file Human_GO_AllPathways_no_GO_iea_April_15_2013_symbo.gmt and click on Ok.
Click on … next to the text box of Phenotype labels.
Make sure Select source file is set to ES_NT.cls.
Select ES12_versus_NT12 and click on Ok.
Click on the down arrow next to the text box for Collapse dataset to gene symbols. Select false.
Click on the down arrow next to the text box for Permutation type. Select gene_set.
Click on Show next to Basic fields.
Click in text box next to Analysis name and rename (example:estrogen_treatment_12hr_gsea_enrichment_results).
Click on … next to “Save results in this folder text box. Navigate to the folder where you wish to save the results (preferably the same directory where all the input files have been saved).
Click on Run in the bottom right corner.
Note: repeat steps 14 - 21 for the 24hrs time-point but use ES24_versus_NT24 phenotype labels in step 14 instead and in step 20 change the Analysis name (example:estrogen_treatment_24hr_gsea_enrichment_results)..
Stpe 2: Generate Enrichment Map with GSEA Output
- Open Cytoscape
- Click on Plugins / Enrichment Maps / Load Enrichment Results
- Make sure the Analysis Type is set to GSEA
- Please select the following files by clicking on the respective (...) button and selecting the file in the Dialog:
NO GMT file is required for DAVID Analysis
Dataset 1 / Expression: Estrogen_expression_file.txt (OPTIONAL)
Dataset 1 / Enrichments: 12hr_David_Output.txt
Click on "Dataset 2
" to expand the panel
Dataset 2 / Expression: leave empty
Dataset 2 / Enrichments 1: 24hr_David_Output.txt (OPTIONAL)
- Tune Parameters
P-value cut-off 0.001
Q-value cut-off 0.05
- Check Overlap Coefficient
Overlap coefficient cut-off 0.6
- Build Enrichment Map
- Go to View, and activate Show Graphics Details
Step 3: Examining Results
Legend:
- Node (inner circle) size corresponds to the number of genes in dataset 1 within the geneset
- Node border (outer circle) size corresponds to the number of genes in dataset 2 within the geneset
- Colour of the node (inner circle) and border(outer circle) corresponds to the significance of the geneset for dataset 1 and dataset 2, respectively.
- Edge size corresponds to the number of genes that overlap between the two connected genesets. Green edges correspond to both datasets when it is the only colour edge. When there are two different edge colours, green corresponds to dataset 1 and blue corresponds to dataset 2.
NOTE: if you are using two enrichment sets you will see two different colours of edges in the enrichment map. When the set of genes in the two datasets are different (for example, when you are comparing two different species or when you are comparing results from two different platforms) the overlaps are computed for each dataset separately as there is a different set of genes that the enrichments were calculated on. In this case, since the enrichments were reduced to only a subset of most differentially expressed at each time point the set of genes the enrichments are calculated on are different and overlap are calculated for each set separately.