Exercise 5 - Task 2

Gene regulation networks / Gene ontology

• Use the Cytoscape BiNGO plugin to find statistically overrepresented GO terms of the "Biological Process" category in the network (Top/Bottom 150 from 5.1.2).
  Hint: install the Bingo plugin first
  
  
  • Which are the 15 most prominent/significant (before correction) "Biological Processes"?
    saved table, can be imported into Excel
  • Show (highlight) 5 selected categories in the network (gene expression)
  • Which pathways could be affected by the PGJ2 treatment?


• Restart Cytoscape and import k-means-up-down.txt (up and down regulated clusters) as Table
  
  Up and down regulated clusters: k-means-up-down.txt


• construct a Expression/Correlation network
• Merge expression table with network
• Use the Cytoscape BiNGO plugin to find statistically overrepresented GO terms of the "Molecular Function" category in the network
• Which genes are where annotated with "transcription factor activity"?

• Map the gene expression data from Wang et al. 2006 to the human p38 MAPK Signaling pathway.
  
  Hint: install the "WikiPathways" plugin first
  
  
  • Import the human p38 MAPK signaling pathway from the WikiPathways database (File -> Import -> Network from public Databases…). Search for human “p38 MAPK Singaling pathway”, which is affected by PGJ2 (see Wang et al. 2006)
    
    
    
  • Map the gene expression data to the pathway
    
    Data File: GDS2104-mean.txt
    
    
    • File -> Import -> Table -> File...
      Attention: define a mapping from "NAME" to "shared name"
    
    
  • display the Node expression values as colors e.g. 24h (remember 5.1.1).
    
  • Make a filter that selects all 2 fold (up or down) regulated genes in any time point, which are in the pathway.
  • Export a list of the genes that are regulated in the pathway


• Map the gene expression data from Wang et al. 2006 to the known human DoRothEA transcription factor - target interactions and find all 1st and 2nd level targets of the DDIT3 transcription factor that are regulated at the timepoint 24h. Remember: our GO anlysis identified the DDIT3 gene as strongly upregulated transcription factor.
  
  Hint: install the "OminPath" plugin first
  
  
  • Import the human DoRothEA TF-target interactions from the OmniPath (TF confidence level A,B)
    
    
  • Map the gene expression data to the network
    
    Data File: GDS2104-mean.txt
    
    
    • File -> Import -> Table -> File...
      Attention: define a mapping from Name to gene_symbol (Key Column for Networks <-> NAME)
    
    Hint: Use filters
    • Filter for "Node: gene_symbol" DDIT3 and build a "Chain" using the "Node Adjacency Transformer" to add adjacent nodes where the edges are outgoing and the expression at 24h is < -log2(1.5) or > log2(1.5)
    • Create a new sub-network from the seleced nodes and edges
  • display the Node expression values at 24h as colors (remember 5.1.1).
    
  • Adjust the Style to display gene_symbols as node names and arrows for interactions.
  • Export a list of these target genes

• Analyze the Wang et al. 2006 expression data (> 2 fold top/bottom 150 from 5.1.2) using DAVID
  
  http://david.ncifcrf.gov
  
  
  • Group the regulated genes according to their functional classification
  • How many functional groups did you identify?
  • Identify the transcription factor group and list the transcription factors
  • Find the group for heat shock response:
    
    
    • Report the genes in that group
    • Find all functionally related genes to this group in your gene list (top/bottom 150)
    • Find all highly and very highly functional related genes in the complete human genome
  
  
  • Run Functional Annotation
  • Which KEGG pathways could you identify?
  • Highlight the genes that are involved in the p53-signaling and the MAPK pathway in a graphic representation of the two pathways.