search

bzhanglab / WebGestaltR

R package for WebGestalt
https://bzhanglab.github.io/WebGestaltR/
34 stars 14 forks source link

Online app and R package do not return the same gene ID output with same parameters #33

Open epertham opened 7 months ago

epertham commented 7 months ago

Dear WebGestaltR authors,

I noticed that when I run the same analysis with the online app and the R package, I don't get the same "job summary" output, for the "entrez ID to functional categories" mapping.

Here is a reproducible code for the R analysis, where I get, "among 18 unique entrezgene IDs, 15 IDs are annotated to the selected functional categories and also in the reference list, which are used for the enrichment analysis" :

interestGene <- c("Rap1gap2","Rcsd1","Ret","Rnf207","Ryr2","Scn4a","Sgk1","Slc16a1","Slc6a4","Sphk1","Spry4","Tcap","Tdgf1","Tlcd4","Tnni3","Trim63","Vrtn","Xirp1","Zbtb20")

res1 <- WebGestaltR(enrichMethod="ORA", organism="mmusculus", enrichDatabase="geneontology_Biological_Process", interestGene=interestGene, interestGeneType="genesymbol", referenceGeneType="genesymbol", referenceSet = "genome", projectName = "run1", minNum = 5, maxNum = 2000, sigMethod = "top" )

While for the online analysis, with the same gene list, I have "among 18 unique entrezgene IDs, 14 IDs are annotated to the selected functional categories and also in the reference list, which are used for the enrichment analysis".

I attach to this issue the 2 corresponding html reports in a zipped archive : html_reports.zip

Am I doing something wrong? Or the mapping process is different between the app and the R package?

Thanks a lot for your help !

Best, Emeline

iblacksand commented 5 months ago

Sorry for the delayed response,

I replied on the google group, but I wanted to message here as well. We can use either platform, whatever you prefer.

The server runs the same R package, so the results should match. It may be due to some difference in parameters, but I don't see any on initial inspection. I will try to find any differences that may have caused the one gene to be lost.

Just to be sure of any potential differences, do you happen to know which version of the package you are running? You can get the version by running packageVersion("WebGestaltR") in a R console.

epertham commented 5 months ago

Dear iblacksand, Thanks a lot for your answer. No worry for the delay. I use the ‘0.4.6’ version of the package. Does it help you? If I can provide any elements useful to reproduce my observation, do not hesitate to ask me. Best, Emeline

iblacksand commented 5 months ago

Thank you for the version number. I am able to replicate the same issue as you. I'll look into where the difference is occurring.

Best, John