Suggestion: More informative "Gene_set" information in custom Enrichr run

[dnjst]

Thank you for the program!

I have the same connection lost error many others are reporting #117 #153 #174. The error message was not revealing that it may be throttling from the website, or connection errors.

I would like to get into contact with Enrichr whether it is an intentional throttling or a server issue.

I downloaded the files from https://maayanlab.cloud/Enrichr/#libraries to .gmt files as you suggested in issue #153 and it works offline as a temporary fix.

However, the results dataframes under the "Gene_set" column now just list a name like "CUSTOM47448322283840", which is understandable because it no longer knows which the gene set is.

Would it be possible to give more informative names in the report?

For example:

gene_sets=["./Enrichr/KEGG_2021_Human.gmt", "./Enrichr/GO_Biological_Process_2021.gmt"]

would result in "KEGG_2021_Human.gmt" and "GO_Biological_process_2021.gmt" as the titles in the results dataframe, and only appending all of the numbers if the filenames are identical, or something?

Gene_set	Term	Overlap	P-value	Adjusted P-value	Odds Ratio	Genes
CUSTOM47448322283840	aerobic electron transport chain (GO:0019646)	6/70	1.03010752590765E-07	3.78221977804328E-05	30.6190878775069	UQCRC1;COX6A1;COX8A;UQCRB;COX7C;COX7B
CUSTOM47448322283840	mitochondrial ATP synthesis coupled electron transport (GO:0042775)	6/71	1.12232040891492E-07	3.78221977804328E-05	30.1896447922938	UQCRC1;COX6A1;COX8A;UQCRB;COX7C;COX7B
CUSTOM47448322283840	platelet aggregation (GO:0070527)	5/36	1.06724547431451E-07	3.78221977804328E-05	49.4551884680813	MYL12A;ACTG1;ACTB;HSPB1;GNAS

turns into

Gene_set	Term	Overlap	P-value	Adjusted P-value	Odds Ratio	Genes
GO_Biological_Process_2021.gmt	aerobic electron transport chain (GO:0019646)	6/70	1.03010752590765E-07	3.78221977804328E-05	30.6190878775069	UQCRC1;COX6A1;COX8A;UQCRB;COX7C;COX7B
GO_Biological_Process_2021.gmt	mitochondrial ATP synthesis coupled electron transport (GO:0042775)	6/71	1.12232040891492E-07	3.78221977804328E-05	30.1896447922938	UQCRC1;COX6A1;COX8A;UQCRB;COX7C;COX7B
GO_Biological_Process_2021.gmt	platelet aggregation (GO:0070527)	5/36	1.06724547431451E-07	3.78221977804328E-05	49.4551884680813	MYL12A;ACTG1;ACTB;HSPB1;GNAS

would be nice.

Also possible would be a way to manually specify the names of the Gene sets:

gene_sets={"KEGG" : "./Enrichr/KEGG_2021_Human.gmt", "GO BP" : "./Enrichr/GO_Biological_Process_2021.gmt"}

but I think that would conflict with the dictionary syntax for passing a single list.

[zqfang]

Hi @dnjst,

Yes, for the default behaivour, same id number is append if filename is the same.

The lastest push is what your expected now.

However, the solution I came up for the dictionary input would stores it's index value.

e.g.

genes_sets=["/path/to/a.gmt",  {'term':[] ...}, "KEGG_2021"]

then, the output Gene_set column will be

a.gmt
gs_ind_1
KEGG_2021

I'll upload a new version to pypi soon

[zqfang]

Fixed in new release v1.0.3