Hi, I found that the results of MetaboAnalyst (current online version) and MetaboAnalystR (v3.2.0) are not consistent?
I am using KEGG IDs for enrichment analysis (KEGG pathways).
I just copy the R codes from the MetaboAnalyst (current online version) of enrichment analysis to local R script, and found that the expected / p / fdr values are largely different?
Generally, the local R scripts generate less significant enrichments with larger p values?
Is this a bug or version issue for my local MetaboAnalystR (v3.2.0) ?
I also found the same issue. The p-value is quite different between MetaboAnalyst5.0 (online) and the local MetaboAnalystR (v3.3.0).
Hello Gluck and fountao,
This is a reproducible issue. The reason for this issue is related to the api calling to our API server (api.metaboanalyst.ca), which has not updated to the latest version. I have fixed it. Please use the latest version of MetaboAnalystR and try again.
Bests,
Zhiqiang
1 Like
Thank you for your timely improvement. By the way, it will also be great if you can add the detailed compound names in the downloadable tsv file for the online version of MetaboAnalyst.