rrvgo

Question

[rbj]

Hi, this is a long shot, but my own reading and attempts and ChatGPT have all failed me. The REVIGO webapp produces some great graphs of reduced GO terms. You can largely reproduce this in R standalone with the bioconductor rrvgo package. However, the webapp graphs allow filtering on a so called dispensability parameter. There is no way to access this automatically in the r package rrvgo, so does anyone know what the actual equation for this parameter is? In the paper listed, it is referenced a few times but only it's usefulness not how it is calculated. Thanks Richard

Revigo paper in Plos

Answer 1

[Matthias Zepper]

I have never used the tool or the web app, but the beauty of open-source software is that one can check this easily in the source code.

The code comment summarises it as follows:

Determines 'dispensability' of each GO term in the list. This is done by recursively eliminating one term from a pair of currently most similar GO terms in the set, and remembering the similarity of the pair as the 'dispensability' of the eliminated term. Which GO term is eliminated from a pair is determined by checking the value of a term's property, if exists (if not, one is removed at random).

The full function comprises lines 341 to 600 in the code.

[rbj]

Wow, thanks that's fabulous Matthias. As you say this is the beauty of open source, I am a grumpy reviewer if people don't do this :-) Of course I had looked, but became overwhelmed by the number of core scripts. I spent most of my time searching in utilities and workers... excuses excuses TLDR Thanks so much!! Richard

[Matthias Zepper]

Don't worry, I myself have often not seen the forest for the trees!