ClueGO uses wrong Identifier

13 views

Skip to first unread message

Enton Funkme

unread,

Mar 21, 2019, 7:43:36 AM3/21/19

to cytoscape-helpdesk

Hello,

I am trying to analyze transcriptome data from zebrafish. I am using EnsemblTranscriptIDs as Identifier. If I selevt "EnsembleTrasncriptID" as Input, no GO terms can be found and no network is created. On the other hand, if I select "# Automatic #", a reasonable analysis is performed and a nice network appears. Still, when I look into the clueGO log, it tells me, clueGO used EnsemblProteinID as Identifier... Why is that? I also tried to convert all my EnsembleTranscriptIDs to EnsembleGeneIDs and when I entered those in ClueGO and chose "EnsembleGeneID" everything was fine... I am confused, what can I do? Are the EnsembleProteinID data trustworthy?

Thanks a lot in advance and warm regards,

Henriette

Bernhard

unread,

Mar 21, 2019, 10:09:32 AM3/21/19

to cytoscape-helpdesk

Hi Henriette,

this seems to be a bug that was fixed some time ago (in the header the ProteinID was flipped with TransciptID). If you update the ensembl files this will be fixed (but you will still see (EnsemblProteinID) "Identifier types used: [EnsemblProteinID, EnsemblTranscriptID]", because the older file with the wrong header is kept and will be recreated in case you delete it because it is part of the initial data provided with the organism downloaded. So just ignore the ProteinID mention). Thanks for reporting this. We will update the organism files.