Perfom a Gene Ontology Analysis from GO terms txt file

Question

[Hamtaro]

Hello everybody.

It's the first time I've been confronted with something like this. After performing a differential analysis using a Tomato genome downloaded from this page, I would like to perform a Gene Enrichment analysis on the differentially expressed genes. However, I have never used a GOterm.txt (from here) as a reference.

Could you tell me how I should do it or a guide/app?

Thanks!

[sansan_96]

Hello,

To perform functional enrichment analysis there are packages such as topGO or clusterProfiler that can work very well. Lately I've been using clusterProfiler for the graphs it comes from, maybe you should take a look at the clusterProfiler manual:

https://yulab-smu.top/biomedical-knowledge-mining-book/clusterprofiler-go.html

If your organism does not have support you may need to use AnnotationHub to retrieve GO terms:

https://guangchuangyu.github.io/cn/2017/07/clusterprofiler-maize/#disqus_thread

[Hamtaro]

Thank u for your reply. The problem is that I used the genome and gene annotation from that webpage, not ensemble or ncbi, so I have to use the Goterm.txt file from that webpage

Answer 1

[sansan_96]

I understand, here is an example with topGO where you need a differential expression table (DESeq2 for example) and a table of GO terms:

library(topGO)
library(ALL)

#DEGs table.
 DE<-read.csv("C:/DE.csv")

#Go terms and ID
 genes2GO<-readMappings(file = "C:/Genes2GO.txt",sep = "\t", IDsep = ";")    

#Column with ID and p-value
GeneList <-DE[,6]       
names(GeneList) <-DE[,1]

#Object GO
 GOdata <-new("topGOdata",
          description = "Simple session", ontology = "BP",                 #BP, MF o CC 
          allGenes = GeneList  ,geneSel = topDiffGenes,   
          nodeSize = 10,                                                
          annot=annFUN.gene2GO,gene2GO = genes2GO  )                     

#Test Fisher o KS
 Fisher  <- runTest(GOdata  , algorithm = "classic", statistic = "fisher")
 KS      <- runTest(GOdata  , algorithm = "classic", statistic = "ks")
 KS_elim <- runTest(GOdata  , algorithm = "elim", statistic = "ks")

#Results
 Genetable<-GenTable(GOdata, classicFisher = Fisher,classicKS = KS, elimKS = KS_elim,orderBy = "elimKS", ranksOf = "classicFisher", topNodes =20)

Review the Bioconductor manual.

[sansan_96]

Your table should look like this (GO terms):

geneid  GOid
Zm00001eb000010 GO:0003690;GO:0003727;GO:0019843;GO:0005739
Zm00001eb000020 GO:0004806;GO:0006511;GO:0065007
Zm00001eb000050 GO:0000723;GO:0006281;GO:0006351;GO:0010556