How to add GO pValue in blast2GO ?

Question

[Farbod]

Hi, I want to use REVIGO TreeMap in order to summarize and visualize the GO terms related to my Differentially Expressed Genes/transcripts (DEGs).

I found the GO term for them using Blast2GO.

But it seems that REVIGO needs "pValue"s, which is not included in the .annot file exported from Blast2GO.

How can I add pValue to my GO terms?

Best

NOTE: Blast2GO results example

TRINITY_DN187708_c5_g2_i7 GO:0005515 diphthine methyltransferase

TRINITY_DN190758_c1_g1_i9 GO:0016307 phosphatidylinositol 4-phosphate 5-kinase type-1 alpha

TRINITY_DN190758_c1_g1_i9 GO:0006897

TRINITY_DN190758_c1_g1_i9 GO:0046854

Answer 1

[Puli Chandramouli Reddy]

Hi,

You need to perform enrichment analysis of GO terms for DEGs over all GO terms to get p-values (can use BINGO). You can use just GO terms without p-values also but REVIGO will not give weightage based on p-values.

[Farbod]

Hi and thank you. would you please explain more what you mean by "over all GO terms" here? as I have collected GO terms just for my 200 DEGs. Thanks

[Puli Chandramouli Reddy]

GO terms to all genes of you model organism.

[Farbod]

Thank you.

I am working on a non model fish. I have 500K transcripts and 200 of them are DEG which I collect GO for them using Blast2GO. The closest model animal to my sample is is zebra fish.

Should I use GO of zebra fish as some kind of background? where can I collect all Danio rerio GO terms?

[Puli Chandramouli Reddy]

Easy way is to map your DEGs to zebrafish genes (find orthologues) and then you can use many of the tools which have zebrafish gene annotations (eg. clusterProfiler)

[Jake Warner]

Perhaps you can use all the mapped GO terms as your background and the GO associated to DEGs to test for enrichment using a hypergeometric or fisher's exact test.

But as Puli said you can use Revigo without p-values.