EnrichmentMap with GSEA shows URLs in the geneset names

[komalsrathi]

I ran GSEA (NOT pre-ranked) on my dataset and here is the structure and parameters of the GSEA run:

structure of GSEA output folder:

human_gsea_control_vs_treat
├── ... (lots of pngs/html)
├── butterfly_plot.png
├── edb
│   ├── gene_sets.gmt
│   ├── human_gsea_control_vs_treat.cls
│   ├── human_gsea_control_vs_treat.rnk
│   └── results.edb
├── ... (lots of pngs/html)
├── gene_set_sizes.xls
├── global_es_histogram.png
├── gsea_report_for_Control_1620063771620.html
├── gsea_report_for_Control_1620063771620.xls
├── gsea_report_for_Treat_1620063771620.html
├── gsea_report_for_Treat_1620063771620.xls
├── ... (lots of pngs/html)
├── heat_map_corr_plot.html
├── human_gsea_control_vs_treat.Gsea.1620063771620.rpt
├── index.html
├── neg_snapshot.html
├── pos_snapshot.html
├── pvalues_vs_nes_plot.png
├── ranked_gene_list_Control_versus_Treat_1620063771620.xls
├── ranked_list_corr_124.png
└── xtools.css

Parameters of GSEA from index.html:

producer_class  xtools.gsea.Gsea
producer_timestamp  1620063771620
param   zip_report  false
param   num 100
param   scoring_scheme  weighted
param   norm    meandiv
param   out /Users/rathik/Projects/Skeletal_Muscle_Patrick/results/human_filtered/gsea_output
param   mode    Max_probe
param   include_only_symbols    true
param   set_max 1500
param   gmx ftp.broadinstitute.org://pub/gsea/gene_sets/h.all.v7.2.symbols.gmt,ftp.broadinstitute.org://pub/gsea/gene_sets/c2.cp.v7.2.symbols.gmt
param   plot_top_x  20
param   nperm   1000
param   order   descending
param   rpt_label   human_gsea_control_vs_treat
param   rnd_seed    42
param   set_min 15
param   res /Users/rathik/Projects/Skeletal_Muscle_Patrick/results/human_filtered/gsea/human_gsea_control_vs_treat.gct
param   create_svgs false
param   cls /Users/rathik/Projects/Skeletal_Muscle_Patrick/results/human_filtered/gsea/human_gsea_control_vs_treat.cls#Control_versus_Treat
param   sort    real
param   create_gcts false
param   help    false
param   save_rnd_lists  false
param   median  false
param   metric  Signal2Noise
param   make_sets   true
param   rnd_type    no_balance
param   gui false
param   permute phenotype
param   collapse    No_Collapse

file    /Users/rathik/Projects/Skeletal_Muscle_Patrick/results/human_filtered/gsea_output/human_gsea_control_vs_treat.Gsea.1620063771620/index.html

I imported this folder in Cytoscape EnrichmentMap and it filled the following fields automatically:

I increased the FDR cut-off to 0.5 and hit Build and got this warning:

I hit Continue to Build and this is what I got:

I am confused why I am getting the full URLs of the gene sets and how do I remove them?

[risserlin]

You are using the GSEA gmt files and EM is optimized for the baderlab genesets that are updated on a monthly basis. (http://download.baderlab.org/EM_Genesets/current_release/)

The easiest fix for the above is just swap the attribute that is being used for the label. It defaults to GS_description but if you change it to Name it should fix it.

In the Style panel expand the Label and change the column from GS_Descr to Name.

[komalsrathi]

Great, thanks for the immediate response!