clusterProfiler enricher question

Question

[skenoy]

I meet a problem when use enricher of clusterProfiler :

my CC background gene number is 9052, print summary(ego), its show BgRatio is 9052, but when i use ego@geneSets to show only 5k, then i use plotGOgraph，the first level of go`s gene number is 5k, not 9052, other levels gene number is right as the same as showing of summary(ego)

1. i use latest version of clusterProfiler

2. library(clusterProfiler)
   annot <- read.table("PFC/annot.C", header=F, sep="\t", quote="", colClasses = "character")
   allg <- read.table("Unigene.fa.bgl", header=F, colClasses = "character")
   deg <- read.table("../CK-VS-T.glist", header=F, colClasses = "character")
   g2g <- annot[, c("V1", "V2")]
   g2t <- annot[, c("V1", "V3")]
   ego <- enricher(deg$V1, universe=allg$V1, pAdjustMethod = "none", minGSSize = 1, TERM2GENE= g2g, TERM2NAME = g2t, pvalueCutoff =1, qvalueCutoff =1)
   write.table(summary(ego)[,-6][,-8], "test/test.C.xls", sep="\t", row.names=F, quote=F)
   ego <- enricher(deg$V1, universe=allg$V1, pAdjustMethod = "none", minGSSize = 1, TERM2GENE= g2g, TERM2NAME = g2t, pvalueCutoff =0.05, qvalueCutoff =1)
   ego@ontology = "CC"
   if(length(rownames(summary(ego))) != 0)
   {
   pdf("test/test.C.pdf")
   if(length(rownames(summary(ego))) > 9)
   {
       plotGOgraph(ego)
   }else{
       plotGOgraph(ego, firstSigNodes = length(rownames(summary(ego))))
   }
   dev.off()
   }

Answer 1

[Guangchuang Yu]

you did not follow the guide, http://stackoverflow.com/questions/5963269/how-to-make-a-great-r-reproducible-example, to make a reproducible example, and I have no idea of your input, output and plot.

BgRatio is 9052, but when i use ego@geneSets to show only 5k

Here, you did not give enough information either.

BgRatio should be something like M/N, M is the number of genes belong the a specific term and N should be the number of all genes in background.

geneSets is a list of geneSet, while each geneSet contains genes that belong to that term, that is corresponding to M.

It seems here 9052 is N and 5k is M and of course they M << N.

[skenoy]

yes, I know BgRatio is format to M/N

9052 is N, the number of geneSets is 5k, are they the same? i think it must be the same, or i use plotGOgraph show the first level of go`s gene number is 5k, not 9052

[Guangchuang Yu]

pls install the newest release of DOSE, it should solve your issue.

Answer 2

[Guangchuang Yu]

reproducible example, pls, otherwise I can't get your idea.

[GenoMax]

Please use ADD REPLY/ADD COMMENT to respond to existing posts. SUBMIT ANSWERS should be reserved for new answers to original question.

[Guangchuang Yu]

OK. thanks for the info