ClueGO Nodes and Edges

[BIOMICS]

Hi all, 

In ClueGO " the size of the nodes is mentioned to correspond to the statistical significance of the terms associated with it. 

I wish to clarify few things are how the size of the node is derived. 

1) The type of statistics used to derive the size of the nodes.  

2) is it right to say the terms in the nodes are overrepresented or enriched for a specific background e.g., for all the human genes. 

2) Does the reference set options (1) selected ontologies reference set (2) predefined reference set, (3) Custom reference set, influence the size of the node. 

3) in the reference set options if I select the option "selected ontologies reference set" please clarify what will be the ontologies that ClueGO will use as the reference set. Does it correspond to all the genes in the human genome? 

4) Does the software allow for no background gene list. If there is an option to select this how could this be fed into the software to perform the analysis? In this case, again does the software take all human genes as the background for calculations. 

Thank you,

Nara

[Bernhard]

Hi Nara, here the answers

On Friday, March 16, 2018 at 6:02:58 AM UTC+1, BIOMICS wrote:

Hi all, 

In ClueGO " the size of the nodes is mentioned to correspond to the statistical significance of the terms associated with it. 

I wish to clarify few things are how the size of the node is derived. 

1) The type of statistics used to derive the size of the nodes.  

As described in the help (click on the (?) icon in the stats section) the pValues are calculated base on the hypergeometric statistics (fisher's excact test). The legends for the node size you get on the (?) icon next to the Visual Style option you select.

2) is it right to say the terms in the nodes are overrepresented or enriched for a specific background e.g., for all the human genes. 

Yes the terms that are significant (based on the FDR correction you selected) have a gene set that is overrepresented (more genes from your input set are found in that term than expected by chance) compared to the not significant terms.

2) Does the reference set options (1) selected ontologies reference set (2) predefined reference set, (3) Custom reference set, influence the size of the node. 

Yes, by default the sum of the unique genes from all selected ontologies are use. See previous answers.

3) in the reference set options if I select the option "selected ontologies reference set" please clarify what will be the ontologies that ClueGO will use as the reference set. Does it correspond to all the genes in the human genome? 

No, it refers to the unique genes from all selected ontologies. If you want the whole genome you can use EntrezGeneIDs as reference.

4) Does the software allow for no background gene list. If there is an option to select this how could this be fed into the software to perform the analysis? In this case, again does the software take all human genes as the background for calculations. 

No, there is no such option.

Thank you,

Nara

Best