ggtree adds unexpected branch/node

[Edouard94]

trafficstars

Hello,

I am struggling with the following code I'm using to annotate my tree. It seems that ggtree adds an extra branch/node to the tree that should not be there, which disrupts all the bootstrap values.

My current code and the tree display are as follows:

# Load libraries
library(ggtree)
library(ggplot2)
library(ape)
library(tidytree)
library(treeio)
library(dplyr)
library(extrafont)
loadfonts()

# Read in your tree file (replace "treefile.nwk" with the path to your tree file)
tree <- read.tree("18S_PutativeSeqs_iqtree.treefile")

# Define the outgroup
outgroup <- "Chromera_velia_strain_CMS22_DQ174731.1"
# Reroot the tree with "Chromera_velia" as the outgroup
tree <- root(tree, outgroup, resolve.root = TRUE)

# Find the index of the tip labels to be renamed
index1 <- which(tree$tip.label == "Cryptosporidium_muris_strain_Calf_genotype_IDVS-811_AF093496.1")
index2 <- which(tree$tip.label == "Ancora_cf._sagittata_isolate_AncoraWSBS2010_KX982504.1")
index3 <- which(tree$tip.label == "Nematopsis_temporariae_isolate_Rd_E8_24L4_KT717659.1")
# Rename the tip labels
tree$tip.label[index1] <- "Cryptosporidium_muris_AF093496.1"
tree$tip.label[index2] <- "Ancora_cf._sagittata__KX982504.1"
tree$tip.label[index3] <- "Nematopsis_temporariae_KT717659.1"

# This shortens your tree to fit tip labels. Adjust the factor for a better fit.
xlim_adj <- max(ggtree(tree)$data$x) * 1.5

# Define the data for geom_hilight
highlight_data <- data.frame(
  node_light = c(111, 180, 154, 169, 179, 150, 151, 142, 134, 120, 122),  # Node IDs
  fill_color = c("#00BC00", "#B66DFF", "#D82632", "#FFCA99", "#999999", "#00AACC", "#00AACC", "#00AACC", "pink", "brown", "#598541"))  # Fill colors

# Plot the tree with new labels
p <- ggtree(tree, ladderize = TRUE) + #ggtree with "ladderize = TRUE" makes the tree look more "balanced" by reordering the nodes
  geom_tiplab(hjust = 0, size = 5, align=TRUE, linesize=.5, offset=0.001, fontface="italic", family="Times New Roman") + #geom_tiplab affects the way the "tips" are displayed (e.g., font, size, color, italics, etc.).
  geom_treescale(y = -0.95, fontsize = 3.9) +
  geom_text2(aes(label = as.numeric(label), 
           subset = !is.na(as.numeric(label)) & as.numeric(label) > 49 & as.numeric(label) <101), 
            vjust = -0.5, hjust = -0.2, size = 3, check_overlap = TRUE) + # geom_text2 affects the text indicating the BS values. Here we choose not to display the values below fifty.
  #labs(title = "18S Phylogenetic tree") +
  theme(legend.text = element_text(size=8)) + # Position of the legend
  xlim(0, xlim_adj*2) +
  #geom_text(aes(label=ifelse(!isTip, node, '')), hjust=-0.5, vjust=0.5, size=3) + # see the nodes number
  geom_cladelabel(node = 111, label = "Terrestrial Gregarines 2", offset = .36, align = TRUE, fontsize = 5, color = "#00BC00", family = "Times New Roman") +
  geom_cladelabel(node = 180, label = "Capitellid Gregarines", offset = .36, align = TRUE, fontsize = 5, color = "#B66DFF", family = "Times New Roman") +
  geom_cladelabel(node = 154, label = "Cephaloidophoroidea", offset = .36, align = TRUE, fontsize = 5, color = "#D82632", family = "Times New Roman") +
  geom_cladelabel(node = 169, label = "Lecudinoidea", offset = .36, align = TRUE, fontsize = 5, color = "#FFCA99", family = "Times New Roman") +
  geom_cladelabel(node = 179, label = "Incertae sedis", offset = .36, align = TRUE, fontsize = 5, color = "#999999", family = "Times New Roman") +
  geom_cladelabel(node = 150, label = "Archigregarine", offset = .36, align = TRUE, fontsize = 5, color = "#00AACC", family = "Times New Roman") +
  geom_cladelabel(node = 151, label = "Archigregarine", offset = .36, align = TRUE, fontsize = 5, color = "#00AACC", family = "Times New Roman") +
  geom_cladelabel(node = 142, label = "Archigregarine", offset = .36, align = TRUE, fontsize = 5, color = "#00AACC", family = "Times New Roman") +
  geom_cladelabel(node = 134, label = "Core Apicomplexa", offset = .36, align = TRUE, fontsize = 5, color = "pink", family = "Times New Roman") +
  geom_cladelabel(node = 120, label = "Cryptosporidium", offset = .36, align = TRUE, fontsize = 5, color = "brown", family = "Times New Roman") +
  geom_cladelabel(node = 122, label = "Actinocephaloidea", offset = .36, align = TRUE, fontsize = 5, color = "#598541", family = "Times New Roman") +
  geom_hilight(data = highlight_data, aes(node = node_light, fill = fill_color), alpha = 0.4, align="both") +
  scale_fill_identity(guide = FALSE)

# Display the tree
p

Part of the tree for better readibility:

The expected bootstrap values:

Any help or insights on why the ggtree function might be causing this issue would be greatly appreciated!

[Edouard94]

I tried running the code by isolating different objects and parameters, but I can't identify the cause of this problem. Other tools, such as FigTree and Presto, display the expected tree correctly. However, ggtree seems to add a branch between Leidyana erratica and Gregarina blattarum, which shifts all the previous bootstrap values, making the tree incorrect.

Please let me know if additional data would help investigate this issue. Meanwhile, I am adjusting all the bootstrap values manually.

Thank you for your assistance.

[brj1]

Your code is shifting the bootstrap values in the y-axis. Try changing vjust = -0.5 to vjust = 0 in geom_text2. You might want to adjust the hjust parameter as well.

  geom_text2(aes(label = as.numeric(label), 
           subset = !is.na(as.numeric(label)) & as.numeric(label) > 49 & as.numeric(label) <101), 
            vjust = 0, hjust = -0.2, size = 3, check_overlap = TRUE)