This is a test version of Biostars. For the public version, visit https://www.biostars.org.
Does the gene universe for enrichGo need to be a list of gene names?

Hello,

When using the enrichGo does the gene universe need to be a gene list or can it be a named list with logfc?

Currently I am doing this:

genes <- names(gene_list)[abs(gene_list)> 0.9]

go_enrich <- enrichGO(gene = genes,
                      universe = names(gene_list),
                      OrgDb = org.Hs.eg.db,
                      keyType="SYMBOL",
                      ont = "ALL",
                      pAdjustMethod = "fdr",
                      pvalueCutoff = 0.05,
                      readable = TRUE)

I am asking this question because I did an entire analysis with the universe being gene_list and not names(gene_list) and it worked and gave me very good results. Then I understood that this was an error and that should be the names(list) resulting in very few significant pathways...

With this I ask why does enrichGo not give an error when you pass gene_list? Does it have a special mode where it takes logfc values into consideration when doing the ORA analysis? Are just names the correct form?

Best Regards!

enrichgo rna-seq overrepresentation ora clusterprofiler
• 2,216 views • link
updated by Basti • Basti • written by Manuel Sokolov Ravasqueira • Manuel Sokolov Ravasqueira

1 answer

Normally if gene_list was not a gene list, there should be a warning message indicating universe is not in character and will be ignored.... So the universe becomes all the genes listed in the internal annotation database.

The correct way would be to use names(gene_list) in your case, because it represents the genes expressed in your dataset

• 0 views • link
written by Basti • Basti

Log in to answer this question.

More posts like this