HierTree

Plotting a hierarchical tree

This script can be used to plot a hierarchical tree like the one presented in Yao et al., 2023 Figure 1. The input can be a data frame with rows representing the leaf-level and columns represent annotations. This data frame needs to be restructured to a parent-child (hierarchical network data frame), which is what the ‘as.hierDF’ does.

require(dplyr)
require(ggplot2)
require(ggnewscale)
require(ggraph)
require(igraph)
require(tidyverse)
#theme_set(theme_void())
require(ggrepel)

Let’s start by loading some dummy hierarchical data.

URL = 'https://allen-brain-cell-atlas.s3-us-west-2.amazonaws.com/metadata/WMB-taxonomy/20231215/cl.df_CCN202307220.xlsx'
data = rio::import_list(URL)

colors <- rio::import("https://allen-brain-cell-atlas.s3-us-west-2.amazonaws.com/metadata/WMB-taxonomy/20231215/views/cluster_to_cluster_annotation_membership_color.csv")

cl.df <- data$cluster_annotation
cl.df <- cl.df[cl.df$class_label != "LQ",]

# add colors to cluster data frame
colors$cluster_alias <- as.character(as.integer(colors$cluster_alias))
cl.df <- cl.df %>% left_join(colors, by=c("cl"="cluster_alias"))

select.columns <- colnames(cl.df)[grep("^supertype", colnames(cl.df))]
st.df <- cl.df %>% group_by_at(select.columns) %>% summarise(n=n())

select.columns <- colnames(cl.df)[grep("^subclass", colnames(cl.df))]
sc.df <- cl.df %>% group_by_at(select.columns) %>% summarise(n=n())

select.columns <- colnames(cl.df)[grep("^class", colnames(cl.df))]
c.df <- cl.df %>% group_by_at(select.columns) %>% summarise(n=n())

A hierarchic structure is basically a set of nodes, with edges linking nodes. Let’s create an edge list for plotting using the package.We’ll do this for a subset of the data.

devtools::source_gist("https://gist.github.com/cvanvelt/e0e48336f01c49aa616aaab4abf997d8")

cl.df <- cl.df[cl.df$class_id %in% c(6,7),]

hierDF <- as.hierDF(cl.df, levels = c("class_id_label", "subclass_id_label","supertype_id_label"),rootname="wmb")

## [1] 1
## [1] 2
## [1] 3

# Create a graph object
graph <- graph_from_data_frame( hierDF)


dummy <- ggraph(graph, layout = 'dendrogram', circular = FALSE) + 
  geom_edge_diagonal()

dend_leaves <- dummy[["data"]] %>% filter(leaf == TRUE)

n_leaves <- nrow(dend_leaves)

dend_leaves <- dend_leaves %>% 
  left_join(st.df[,c("supertype_id_label", "supertype_color","supertype_id")], by=c("name"="supertype_id_label"))

subclass.df <- dummy[["data"]] %>% 
  filter(name %in% sc.df$subclass_id_label) %>% 
  left_join(sc.df[,c("subclass_id_label","subclass_id", "subclass_color")], by=c("name"="subclass_id_label"))
subclass.df$subclass_id <- gsub( " .*$", "", subclass.df$name) 

class.df <-  dummy[["data"]] %>% 
  filter(name %in% c.df$class_id_label) %>% 
  left_join(c.df[,c("class_id_label", "class_color")], by=c("name"="class_id_label"))

Next use ggraph to plot the ‘dendrogram’ and add additional layers of labeling using standard ggplot.

flat_plot = ggraph(graph, layout = 'dendrogram', circular = FALSE) + 
  geom_edge_diagonal(width = 0.25,
                     color="grey50") +
  #supertype
  geom_point(data = dend_leaves, 
             aes(x=x, y=y, color=supertype_color),
             cex=1,
             shape=19) +
  scale_color_identity(guide = "none") +
  # subclass
  new_scale_color() +
  geom_point(data=subclass.df, 
             aes(x=x, y=y, color= subclass_color),
             cex=2,
             shape=19)+
  geom_text(data=subclass.df, 
             aes(x=x, y=y, label= subclass_id),
             size=3,hjust=0, vjust=0.5,
             angle=90)+ #,             direction='y')+
    scale_color_identity(guide = "none") +
  # class
  new_scale_color() +
  geom_point(data=class.df, 
             aes(x=x, y=y, color= class_color),
             cex=2,
             shape=19)+
  geom_text_repel(data=class.df, 
             aes(x=x, y=y, label= name),
             size=3,hjust=1, vjust=0.5,
             direction='y')+
  scale_color_identity(guide = "none") +
  # add other levels if needed
  geom_text(data=dend_leaves,
            aes(x = x,
                y = y-0.1,
                label = name),
            angle = 90,
            hjust = 1.0,
            vjust = 0.5,
            size = 3,
            lineheight=0.1) +
  scale_x_continuous(limits = c(-1,n_leaves + 1),
                     expand=c(0,0)) +  
  coord_cartesian(clip = 'off')  +
  theme_void()  +
  theme(plot.margin = margin(t = 0, r = 0, b = 120, l = 0,))

flat_plot