Hierarchical edge bundling plots helps visualizing correlation or similarity between clusters

Source: R/Plots.R

Hierarchical edge bundling plots helps visualizing correlation or similarity between clusters

edge_bundling_GSEA(
  SeuratObj,
  by = "GO",
  link_threshold = 0.6,
  link_width = 0.9,
  p.adjust_thresh = 0.05,
  method = "cosine similarity",
  node.by = "cluster",
  group.by = "cellType"
)

Arguments

SeuratObj

Seurat object
by

GO KEGG Reactome MSigDb WikiPathways DO NCG DGN.
link_threshold

only show links of similarity or correlation above threshold
link_width

width of links
p.adjust_thresh

threshold of p.adjust to filter pathways used to calculate cosine similarity/correlation
method

one of c('cosine similarity', 'pearson', 'spearman')
node.by

draw each node as node.by, column of SeuratObj@meta.data
group.by

color each node by group.by, column of SeuratObj@meta.data

Examples

if (FALSE) {
edge_bundling_GSEA(SeuratObj, link_threshold=0.6, p.adjust_thresh=0.05,
             method='cosine similarity', node.by='cluster', group.by='cellType')
edge_bundling_GSEA(SeuratObj, link_threshold=0.6, p.adjust_thresh=0.05,
             method='pearson', node.by='cluster', group.by='cellType')
}