Representative = null in REVIGO Treemap output

[maglund]

Hi,

how should I interpret/handle rows where the representative column in the REVIGO Treemap output is null - should I simply copy what its name column says? Seems reasonable as the row below has this name as representative (included a short example below).

And I understand that this is more of a REVIGO question, but what does representative = null signify?

Thank you for a nice visualization tool!

Name	Frequency	Value	Uniqueness	Dispensability	Representative
response to abiotic stimulus	0.584	-4.000	0.955	0.000	null
cellular response to oxygen levels	0.022	-4.000	0.935	0.264	response to abiotic stimulus
response to interleukin-12	0.001	-4.000	0.954	0.335	response to abiotic stimulus
response to cadmium ion	0.019	-4.000	0.948	0.388	response to abiotic stimulus
response to interleukin-1	0.017	-3.699	0.947	0.639	response to abiotic stimulus

[kubu4]

Also wondering about this issue, particularly since the "null" entry gets incorporated into the CirGO plot, which is less than ideal:

[kubu4]

Well, it's definitely a REVIGO thing. Here's an update from their FAQ:

Q: When I export the TreeMap to CSV, what does NULL (-1) value in "Representative" column means?
A: The column "Representative" tells which terms are classified as similar, and one term which is marked as "Representative" of that group which has a NULL (-1) reference (the term represents itself). In previous versions of Revigo the column "Representative" was filled with representative ID, but it was more accurate to mark it as NULL because that GO term doesn't have representative, it is the "Representative".

It would be amazing if CirGO checked for these null entries and replaced them with the appropriate GO term when it converts the REVIGO TreeMap csv.

[pakmikeduffy]

I really like the CirGO visualisation, it would be great if it could be updated to address this issue

[kubu4]

@pakmikeduffy - I submitted a pull request to fix this awhile ago. My recommendation would be to fork the repo and incorporate the changes in the Pull Request. They're very simple and just require adding one line to each of the three conversion files.

Here's what the Pull Request looks like, if you're interested in trying to fix this yourself:

40af52d