The enrichplot package implements several visualization methods to help interpreting enrichment results. It supports visualizing enrichment results obtained from DOSE (Yu et al. 2015), clusterProfiler (Yu et al. 2012), ReactomePA (Yu and He 2016) and meshes. Both over representation analysis (ORA) and gene set enrichment analysis (GSEA) are supported.
Enrichment Analysis
Over Representation Analysis
Gene Set Enrichment Analysis
Visualization methods
Bar plot
Bar plot is the most widely used method to visualize enriched terms. It depicts the enrichment scores (e.g. p values) and gene count or ratio as bar height and color.
Dot plot
Dot plot is similar to bar plot with the capability to encode another score as dot size.
p1 <- dotplot(edo, showCategory=30) + ggtitle("dotplot for ORA")
p2 <- dotplot(edo2, showCategory=30) + ggtitle("dotplot for GSEA")
plot_grid(p1, p2, ncol=2)
Users can use formula to specify derived variable of x-axis.
## [1] 17381
Gene-Concept Network
Both the barplot and dotplot only displayed most significant enriched terms, while users may want to know which genes are involved in these significant terms. The cnetplot depicts the linkages of genes and biological concepts (e.g. GO terms or KEGG pathways) as a network. GSEA result is also supported with only core enriched genes displayed.
## convert gene ID to Symbol
edox <- setReadable(edo, 'org.Hs.eg.db', 'ENTREZID')
cnetplot(edox, foldChange=geneList)
UpSet Plot
The upsetplot is an alternative to cnetplot for visualizing the complex association between genes and gene sets. It emphasizes the gene overlapping among different gene sets.
Heatmap-like functional classification
The heatplot is similar to cnetplot, while displaying the relationships as a heatmap. The gene-concept network may become too complicated if user want to show a large number significant terms. The heatplot can simplify the result and more easy to identify expression patterns.
Enrichment Map
Enrichment map organizes enriched terms into a network with edges connecting overlapping gene sets. In this way, mutually overlapping gene sets are tend to cluster together, making it easy to identify functional module.
ridgeline plot for expression distribution of GSEA result
The ridgeplot will visualize expression distributions of core enriched genes for GSEA enriched categories. It helps users to interpret up/down-regulated pathways.
running score and preranked list of GSEA result
Running score and preranked list are traditional methods for visualizing GSEA result. The enrichplot package supports both of them to visualize the distribution of the gene set and the enrichment score.
Another method to plot GSEA result is the gseaplot2 function:
The gseaplot2 also supports multile gene sets to be displayed on the same figure:
User can also displaying the pvalue table on the plot via pvalue_table parameter:
gseaplot2(edo2, geneSetID = 1:3, pvalue_table = TRUE,
color = c("#E495A5", "#86B875", "#7DB0DD"), ES_geom = "dot")
User can specify subplots to only display a subset of plots:
The gsearank function plot the ranked list of genes belong to the specific gene set.
Multiple gene sets can be aligned using cowplot:
pp <- lapply(1:3, function(i) {
anno <- edo2[i, c("NES", "pvalue", "p.adjust")]
lab <- paste0(names(anno), "=", round(anno, 3), collapse="\n")
gsearank(edo2, i, edo2[i, 2]) + xlab(NULL) +ylab(NULL) +
annotate("text", 0, edo2[i, "enrichmentScore"] * .9, label = lab, hjust=0, vjust=0)
})
plot_grid(plotlist=pp, ncol=1)
pubmed trend of enriched terms
One of the problem of enrichment analysis is to find pathways for further investigation. Here, we provide pmcplot function to plot the number/proportion of publications trend based on the query result from PubMed Central. Of course, users can use pmcplot in other scenarios. All text that can be queried on PMC is valid as input of pmcplot.
terms <- edo$Description[1:3]
p <- pmcplot(terms, 2010:2017)
p2 <- pmcplot(terms, 2010:2017, proportion=FALSE)
plot_grid(p, p2, ncol=2)
References
Yu, Guangchuang, and Qing-Yu He. 2016. “ReactomePA: An R/Bioconductor Package for Reactome Pathway Analysis and Visualization.” Molecular BioSystems 12 (2):477–79. https://doi.org/10.1039/C5MB00663E.
Yu, Guangchuang, Li-Gen Wang, Yanyan Han, and Qing-Yu He. 2012. “clusterProfiler: an R Package for Comparing Biological Themes Among Gene Clusters.” OMICS: A Journal of Integrative Biology 16 (5):284–87. https://doi.org/10.1089/omi.2011.0118.
Yu, Guangchuang, Li-Gen Wang, Guang-Rong Yan, and Qing-Yu He. 2015. “DOSE: An R/Bioconductor Package for Disease Ontology Semantic and Enrichment Analysis.” Bioinformatics 31 (4):608–9. https://doi.org/10.1093/bioinformatics/btu684.