When using GOseq on the Galaxy platform, the resulting dot plot placed two distinct GO terms on the same row because they shared a similar beginning in their descriptions. Specifically, these terms were: GO:0016635 – oxidoreductase activity, acting on the CH-CH group of donors, quinone or related compound as acceptor
and GO:0016627 – oxidoreductase activity, acting on the CH-CH group of donors.
I think this overlap likely occurred because Galaxy auto-generates plot labels based on truncated or partial GO descriptions, if so, Is there any method to fix this issue or to regenerate the dotplot starting from the output of the Goseq (table)! Thanks in advance for any help1
This is interesting, I think your guess about what is going on sounds correct, and I bet this is happening in the underlying Bioconductor implementation of the tool too, since that is what is running in Galaxy (including the reference data) and I can find it reported as happening in related functions (topGO: how to rettrieve (long) GO term description without abbreviating).
I don’t have a good answer for you but I am wondering if the Bioconductor team might. If they provide you with R code, you can move your data from your history over into an R environment, run the operations, then move the result back into your history. How to → Hands-on: RStudio in Galaxy / RStudio in Galaxy / Using Galaxy and Managing your Data
Not a great answer, but if you are able to solve this, it would be great if you wanted to post back the solution for others (and maybe you have already been able to do that?).
