YuLab-SMU / clusterProfiler

:bar_chart: A universal enrichment tool for interpreting omics data
https://yulab-smu.top/biomedical-knowledge-mining-book/
965 stars 246 forks source link

gseGO Produces Results Table with Zero Rows #702

Open DarioS opened 3 days ago

DarioS commented 3 days ago

Enrichment analysis is empty.

> head(geneList)
 NQO1 PAPPA MGST1  VCAN  CCN2 CCND2 
 5.36  3.17  2.99  2.96  2.35  2.26
> result <- gseGO(gene = geneList, ont = "BP", keyType = "SYMBOL", OrgDb = "org.Hs.eg.db")
> dim(result@result)
  0 8

but overrepresentation analysis works on the same input data.

> result <- enrichGO(gene = sample(names(geneList), 20), ont = "BP", keyType = "SYMBOL", OrgDb = "org.Hs.eg.db")
> head(result@result)[, 1:5]
                   ID                            Description GeneRatio   BgRatio         pvalue
GO:0051216 GO:0051216                  cartilage development      5/19 209/18888 0.000001620682
GO:0061448 GO:0061448          connective tissue development      5/19 288/18888 0.000007768814
GO:0031589 GO:0031589                cell-substrate adhesion      5/19 359/18888 0.000022524359
GO:0070301 GO:0070301 cellular response to hydrogen peroxide      3/19  67/18888 0.000039691721
GO:0006809 GO:0006809      nitric oxide biosynthetic process      3/19  77/18888 0.000060224472
GO:0046209 GO:0046209         nitric oxide metabolic process      3/19  84/18888 0.000078099162

Does enrichment analysis not work?

guidohooiveld commented 3 days ago

What happens if you set within the function gseGO the argumentpvalueCutoff = 1 (instead of the default value of 0.05)?

DarioS commented 3 days ago

Ah, then there are rows of results. However, I would expect there to always be results, such as for limma's topTable output.