Here is what I am doing when clusterProfiler did not give any results with calculated universe genes:

res.cer = as.data.frame(results(dds_cer_comb))

prep_genelist <- function(res){ overallBaseMean <- as.matrix(res[, "baseMean", drop = F]) de_genes = rownames(res[which(res$padj < 0.05), ])

library(genefilter) backG <- genefinder(overallBaseMean, de_genes, 10, method = "manhattan") backG <- rownames(overallBaseMean)[as.vector(sapply(backG, function(x)x$indices))] backG <- setdiff(backG, de_genes) genelist = list(de_genes, backG) names(genelist) = c("de_genes", "backG_genes") return(genelist) }

cer.genes = prep_genelist(res.cer) resultGO = enrichGO(cer.genes$de_genes, universe = cer.genes$backG_genes, keytype = "ENSEMBL", qvalueCutoff = 0.2, ont="BP", minGSSize = 5, maxGSSize = 5000, OrgDb='org.Hs.eg.db')

resultGO #

over-representation test

#

...@organism Homo sapiens

...@ontology BP

...@keytype ENSEMBL

...@gene chr [1:4257] "ENSG00000000003" "ENSG00000000419" "ENSG00000001167" "ENSG00000001497" ...

...pvalues adjusted by 'BH' with cutoff <0.05

...0 enriched terms found

'data.frame': 0 obs. of 9 variables: $ ID : chr $ Description: chr $ GeneRatio : chr $ BgRatio : chr $ pvalue : num $ p.adjust : num $ qvalue : num $ geneID : chr $ Count : int

...Citation

Guangchuang Yu, Li-Gen Wang, Yanyan Han and Qing-Yu He. clusterProfiler: an R package for comparing biological themes among gene clusters. OMICS: A Journal of Integrative Biology 2012, 16(5):284-287

Unfortunately, clusterProfiler did give any results when I used background (universe) genes which I used for goseq analysis.

Any reproducible example?

compareCluster analysis gave some interesting results but I want to know if it is possible to perform and plot compareCluster type of analysis from the results obtained using GOseq?

This is possible, see here for an example published in BMC Genomics that integrate DAVID with clusterProfiler.

I checked the following

length(cer.genes$de_genes)

[1] 4257

length(cer.genes$backG_genes)

[1] 10605

class(cer.genes$de_genes)

[1] "character"

class(cer.genes$backG_genes)

[1] "character"

Despite of this I tried the following:

resultGO = enrichGO(as.character(cer.genes$de_genes), universe = as.character(cer.genes$backG_genes), keytype = "ENSEMBL", qvalueCutoff = 0.2, ont="BP", minGSSize = 5, maxGSSize = 5000, OrgDb='org.Hs.eg.db')

again no result like before. Can you please point out where exactly I did the wrong? Does universe length has something to do with this?Thanks