To demonstrate the selection of genes for further research, we will be be using Cytoscape (Version 3.6) for the network analysis of upregulated genes in a publically availabe dataset from Gene Expression Omnibus (GEO). In the current dataset (GSE4745) microarrays were used to study the ventricle gene expression changes that underly development of diabetic cardiomyopathy. The microarray data was analysed by GEO2R tool available on the dataset page.
Click on the “Analyze with GEO2R” button.
Clicking on GEO2R brings up the group selection page. In our analysis, we compared the STZ 3-day rat with the control group. Click on the drop down menu to make two groups (Test and Control in this case). Select all the control groups and click on the “Control” in the drop down menu. Then do the same for “Test”.
Once samples have been added to the appropriate groups, click on “Top 250” button.
After a brief analysis, we get the top 250 upregulated and downregulated genes. Click on “save all results” to save the results locally.
The results may open in a browser tab. Click CLTR+S to save the result in a text file on your computer. You may append “.csv” to the filename and open the file in Microsoft excel or Libreoffice calc. While importing this file in excel or calc, choose separator as “Tab” and text delimiter as ” ” “.
Once the file has opened in excel or libreoffice calc, sort the rows as per the logFC value. This is the log of Fold change and it is the value we are interested in. “P.Value” should stay below 0.05. A logFC of 1 signifies the doubling of expression. We selected the genes with logFC > 1.5 and p <0.05.
The selected genes were further analyzed in Cytoscape version 3.6. We used the Bisogenet plugin available in the App manager in Cytoscape to download the Protein-protein interactions of the selected genes.
Paste the names of the genes (from gene.symbol column) in to the “input identifiers” box. Select the organism and map for both genes and proteins. Unselect the microRNA silencing interactions from the Data settings tab. Click “Submit”. Bisogenet populates the main cytoscape window with the interactions of the input genes.
Click the CLTR key and select the isolated nodes (genes are the nodes) which are lying alone or are connected to just one another gene. Press delete to remove these nodes from further analysis. Default Bisogenet settings selects up to 1 neighbour of the input nodes so that even genes not entered initially may be a part of the network.
Once the network has been generated, remove duplicated edges and self loops.
Edit > Remove Self loopsEdit >Remove Duplicated edges
We may now analyze this network by:
Tools > Network Analyzer > Network Analysis > Analyze network
Generate styles for this network by using the calculated Betweenness centrality.
Tools > Network Analyzer > Network Analysis > Generate style from Network
Select “Map Node Size to” and “Node Color to” Betweenness centrality in the dropdown menu.
We will get the following graphics. Selecting any node in the window brings it up in the table menu.
From among the genes with high Betweenness Centrality, we may select genes which may be potential drug targets. The gene druggability may be checked from this site.