Loading gene set collections...
Enrich - a Gene Set Enrichment Analysis (GSEA) tool
Green Listed Correlate Visualize Wermeling Lab How to use How to interpret v.42.2

How to use Enrich

Enrich runs Gene Set Enrichment Analysis (GSEA) entirely in your browser. No data is uploaded to any server.

1. Prepare your data

Your input file should be a .csv or .tsv with at least two columns:

Save your spreadsheet as CSV (comma-separated) or TSV (tab-separated) from Excel, Google Sheets, or R. Ensure the first row contains column headers.

2. Upload and configure

  1. Click Choose file or use Load Example Data to try with sample data
  2. Select which column contains gene names and which contains the ranking metric
  3. Genes will be automatically sorted from highest to lowest metric value

3. Choose gene set collections

Select one or more MSigDB collections (Hallmark is recommended to start). You can also upload a custom GMT file.

4. Run the analysis

Click Run GSEA. The analysis runs in your browser using a Web Worker. With Hallmark (50 sets) this takes seconds; larger collections take longer.

5. Interpret the results

Key statistics

Gene set collection strategy

MSigDB contains over 33,000 gene sets across 7 collections. Running all at once is possible but takes several minutes and produces strict FDR corrections. A recommended approach:

  1. Start with Hallmark (50 sets) — well-defined biological pathways, fast and highly interpretable
  2. Explore specific collections — if Hallmark shows inflammatory signals, try C7 (Immunologic); for metabolic hits, try C2 (Curated: KEGG/Reactome)
  3. Use the Gene Set Browser — search for specific pathways or genes of interest across all collections
  4. Iterative permutation strategy — for large analyses, you can run with lower permutations (e.g. 200) for faster screening, then filter the results table and use “Re-run filtered” with ≥1000 permutations. Note: low permutations give coarser p-values, so borderline hits (FDR ~0.2–0.3) may be missed in the initial screen
  5. Use the Data Type selector — set to "CRISPR / Perturbation" if your data is from a genetic screen or perturbation experiment, for correct interpretation
  6. For large analyses — if your computer struggles with many gene sets, use the “Download R Script” option in the run dialog. This generates a self-contained fgsea script you can run in R — then upload the results back here for visualization

Note on naming

Enrich is not the same tool as Enrichr (Ma’ayan Lab). Enrichr performs over-representation analysis (ORA) on gene lists. Enrich performs Gene Set Enrichment Analysis (GSEA) on pre-ranked gene lists, which considers all genes in the ranking — not just a cutoff-based subset.

How to Interpret Enrichment Data

GSEA walks down your ranked gene list. When pathway genes cluster at the top or bottom, the enrichment score (ES) rises or falls. Below are three example outcomes.

Positive enrichment (NES > 0)
← Peak
Expression: Pathway is upregulated — genes are enriched among those with high fold change
CRISPR: Pathway genes confer resistance — knockouts are enriched among positively selected hits
Negative enrichment (NES < 0)
Trough →
Expression: Pathway is downregulated — genes are enriched among those with low fold change
CRISPR: Pathway genes confer sensitivity — knockouts are enriched among depleted hits
No enrichment (NES ≈ 0)
Expression: No directional change — pathway genes are randomly distributed
CRISPR: No selective effect — pathway knockouts show no consistent phenotype

What about curves with both a positive and negative peak?

+ peak
− peak

Some curves show both a positive and negative deflection. This means the pathway has genes on both ends of the ranked list — some are upregulated/enriched and others are downregulated/depleted. The reported ES is the peak with the larger absolute value. While the pathway is still significantly enriched in one direction, the dual-peak pattern suggests the pathway contains functionally diverse genes with opposing responses to the condition.

Key Statistics

1. Upload Data
CSV or TSV with a gene name column and a numeric ranking column
or try example data
DepMap cancer cell line data
A375Skin (Melanoma)
A549Lung (Adenocarcinoma)
HT29Colorectal (Colon)
RajiBlood (B-cell Lymphoma)
U251Brain (Glioblastoma)
Column settings
Genes will be ranked in descending order by this value
Look for enrich_fgsea_results_*.json in your Downloads folder
2. Gene Set Collections
Well-defined biological states and processes
C2: Curated Pathways
C5: Gene Ontology
Transcription factor and miRNA targets
Oncogenic signature gene sets
Immunologic signature gene sets
Cell type signature gene sets
Browse & select individual gene sets →
Hallmark: 50 gene sets loaded
▶ Upload custom GMT file
3. Analysis Parameters ? Quick guide: Default settings work well for most analyses. Use 1000 permutations for publication results. Increase to 10,000 for very precise p-values. The weight parameter p=1 is the standard GSEA weighting.
4. Run Analysis
📊

Run GSEA to see the enrichment overview and ranked gene list

📈

Run GSEA, then click a gene set in the overview plot or results table to see its enrichment profile here

🧬

Run GSEA to discover shared genes between enriched pathways

New to Enrich?
Load test data for skin, lung, colorectal, brain or blood cancer — or upload your own data set — and click Run GSEA
(both in the sidebar on the left)
The Results Table shows all tested gene sets. Click any row to see its Enrichment Plot. The Overview tab shows a lollipop plot and ranked gene list. Use Overlap to find redundant pathways.
Performance note: Over 33,000 MSigDB gene sets are available, but large analyses may be very slow or become unresponsive depending on your computer. For heavy analyses, Enrich can generate an R script (fgsea) — run it locally in R and upload the results back here to use Enrich’s visualization tools.
GSEA Web Tool — Wermeling Lab, Karolinska Institutet  |  Gene sets from MSigDB v2023.2
āš™ Global Settings

Typography

Export

šŸ’” Double-click text on any graph to edit it
šŸ’” Per-graph settings: click āš™ on each graph header

Gene Set Browser

0 selected

Gene Set Details

Hover over any gene set to see its details here.