tools-only Enrichment Visualization - Usage Guide
Overview
The enrichplot package provides visualization functions for clusterProfiler results, including dot plots, bar plots, networks, and GSEA-specific plots.
Prerequisites
if (!require('BiocManager', quietly = TRUE))
install.packages('BiocManager')
BiocManager::install(c('clusterProfiler', 'enrichplot'))
Quick Start
Tell your AI agent what you want to do:
- "Create a dotplot of my GO enrichment results"
- "Make a gene-concept network from my KEGG enrichment"
- "Show a GSEA running score plot for my top pathway"
Example Prompts
Dot Plots
"Create a dotplot showing the top 20 enriched GO terms"
"Make a dotplot of my KEGG results with adjusted font size for long pathway names"
Network Plots
"Create a gene-concept network colored by fold change"
"Generate an enrichment map clustering similar GO terms together"
GSEA Plots
"Show a GSEA running score plot for the top 3 enriched pathways"
"Create a ridge plot showing fold change distributions for each gene set"
Customization
"Save the enrichment dotplot as a PDF at publication quality"
"Change the color scale to viridis on my enrichment plot"
Comparison Plots
"Create a dotplot comparing enrichment results between up and down regulated genes"
What the Agent Will Do
- Take enrichment results from GO, KEGG, or GSEA analysis
- Select appropriate plot type based on request
- For emapplot, compute pairwise term similarity first
- Generate publication-quality figure with appropriate sizing
- Save to PDF or PNG with specified dimensions
Plot Types Quick Reference
| Plot Type | Function | Best For |
|---|---|---|
| Dot plot | dotplot() | Overview of top terms |
| Bar plot | barplot() | Simple count/ratio display |
| Network | cnetplot() | Gene-concept relationships |
| Map | emapplot() | Term similarity clusters |
| Tree | treeplot() | Hierarchical term grouping |
| Upset | upsetplot() | Overlapping genes |
| GSEA | gseaplot2() | Running enrichment score |
| Ridge | ridgeplot() | Fold change distribution |
| Heatmap | heatplot() | Gene-concept matrix |
Choosing a Visualization
For Over-Representation Results
- Starting point: dotplot() - best overview
- Show relationships: cnetplot() - genes and terms
- Cluster terms: emapplot() - similar terms grouped
- Compare groups: dotplot() with compareCluster result
For GSEA Results
- Starting point: ridgeplot() - all gene sets
- Detailed view: gseaplot2() - running score for specific sets
- Overview: dotplot() - works for GSEA too
Tips
- dotplot() is the most common choice for publications - start there
- For emapplot(), always run pairwise_termsim() first on your enrichment result
- Use showCategory parameter to control number of terms displayed
- For long term names, increase plot width and reduce font size
- cnetplot() with circular = TRUE helps with crowded networks
- gseaplot2() accepts multiple geneSetID values to compare pathways
- Use ggsave() for better control over output dimensions and resolution