search

MaayanLab / enrichr_issues

5 stars 3 forks source link

[BUG] Different gmt files are used in Enrichr server and download package #20

Closed hsiaoyi0504 closed 3 years ago

hsiaoyi0504 commented 3 years ago

It is originally reported to https://github.com/zqfang/GSEApy/issues/122. However, it seems to me that the root cause of the issue is probably because the gmt files that we can download from https://maayanlab.cloud/Enrichr/#stats are different from those used in the Enrichr server.

maxim-k commented 3 years ago

@hsiaoyi0504 data is the same, as gmt files on #stats are generated from the database every time you request them. We didn't develop GSEApy package, so I don't know how does 'local' mode of GSEApy work, but from what I see they calculate values in local mode themselves. That's probably the reason whey values are close, but not identical.

hsiaoyi0504 commented 3 years ago

@maxim-k Thanks for replying to me. However, as I mentioned in the original issue:

One thing I notice is that the total numbers of genes in PI3K-Akt signaling pathway term (354 and 352) are different in these two modes. Does that mean the gmt files used by the Enrichr sever are different from those we can download from Enrichr sever?

Do you mean that the gmt file of KEGG_2019_HUMAN is probably updated recently? I downloaded the gmt file earlier today.

hsiaoyi0504 commented 3 years ago

Oh, I checked with the Malaria term. It seems like the GSEApy might have a bug of parsing gmt file.