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

[Emeline Perthame]

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.

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

[John Elizarraras]

Sorry for the delayed response,

Thank you for the report. The server is running the R package, so the results should match.

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.

Best,

John

[Thanh Thanh Nguyen]

Hi,

I am experiencing exactly the same problem as Emeline. My packageVersion("WebGestaltR") is 0.4.6. Can you please advise?

[Thanh Thanh Nguyen]

By R package, I got 164 out of 240 unique IDs mapped to the custom enrichment category. By web, only 98 out of 240. This significantly skews the enrichment results.

On Friday, January 12, 2024 at 2:36:51 PM UTC-5 John Elizarraras wrote:

[Kim L]

Hi

I am using the online version and I am also getting a different output from what I got two weeks ago. Mainly my top ten groups on the enrichment ratio plot differ significantly in name and ratio score. 

Thank you, 

Kimberly Louisor

[John Elizarraras]

Hello Kimberly,

This is expected, as on May 29, we released WebGestalt 2024, which included updates to the underlying data while also adding other new features. This will change your results, sometimes drastically. The previous version is online at https://2019.webgestalt.org/ if you would like to maintain the same version. You should get the same results at that website. More information about the data update can be found here: https://www.webgestalt.org/docs/version_update.html.

Best,

John

[John Elizarraras]

This is a drastic difference. I previously spent some time looking into the underlying cause of the previous issue, but could not find a source at that time.

Do you mind sharing the list of IDs you are trying to map? The difference is larger, so it may be easier to spot the source of the issue. Also, we recently updated WebGestalt so this issue may have been inadvertently fixed. I will also check the previous example for this, but the update may resolve your problem.

Best,

John