How to adjust the branch length of dendrogram converted from Newick file? - r

Question

Hi I aim to make a heat map of species abundance and also show phylogeny tree in parallel. The phylogeny tree was made in MEGA and input to R as Newick file prior to converting to dendrogram object.
However, the branch length of phylogeny tree in the output figure was not the same length. Can I ask is there any way to make the branch length similar?
I attached the code and the output below. Thank you in advance.
library(DECIPHER)
library(dendextend)
library(ape)
#Input the tree from newick file from Mega and convert into dendrogram
dend <- ReadDendrogram("VPstrainand3-major_ASV_5%_withlength.nwk")
#Make the plot using data from the excel stored in the directory
#load the data
library(openxlsx)
library(dplyr)
heatmap.dataframe <- read.xlsx("Complete.xlsx","Modified", rowNames =TRUE)
heatmap.mat <- as.matrix(heatmap.dataframe)
whitered <- colorRampPalette(c("white", "red"), space = "rgb")(100)
#plot
library(repr)
options (repr.plot.width=20, repr.plot.height=20)
heatmap <- heatmap(as.matrix(heatmap.mat), Rowv = dend, Colv = NA, col = whitered)
print(heatmap)
Output heat map with phylogeny tree

Answer 1

We can do this with a combination of dendrapply() and a simple function - named ForceUltrametric in this example. This form can be used for a couple different tasks, the example present in dendrapply's help file is for changing leaf colors, but you could use this for adding points below leaves, or adjusting labels as well.
A reproducible example of your entire process in R, since we don't have access to your newick file:
library(DECIPHER)
# using one of DECIPHER's built in examples:
db <- system.file("extdata",
"Bacteria_175seqs.sqlite",
package="DECIPHER")
dna <- SearchDB(db,
remove="all")
alignedDNA <- AlignSeqs(dna)
Dist <- DistanceMatrix(myXStringSet = alignedDNA,
includeTerminalGaps = TRUE, # global identity
verbose = TRUE)
tree1 <- IdClusters(myDistMatrix = Dist,
method = "NJ", # a non-ultrametric tree
verbose = TRUE,
showPlot = FALSE, # you can return a plot as the function completes, or not
type = "dendrogram") # return a dendrogram
ForceUltrametric <- function(n) {
if (is.leaf(n)) {
# if object is a leaf, adjust height attribute
attr(n, "height") <- 0L
}
return(n)
}
tree2 <- dendrapply(X = tree1,
FUN = function(x) ForceUltrametric(x))
Using some newick file, which we don't have access to, and cannot guarantee working:
library(DECIPHER)
tree1 <- ReadDendrogram("<yourfilehere>")
ForceUltrametric <- function(n) {
if (is.leaf(n)) {
# if object is a leaf, adjust height attribute
attr(n, "height") <- 0L
}
return(n)
}
tree2 <- dendrapply(X = tree1,
FUN = function(x) ForceUltrametric(x))
It should be noted that while it's pleasant to force dendrograms to be ultrametric in cases where you want to plot data below the leaves, it isn't really an accurate representation of the data if the tree was built with a non-ultrametric method.
tree1
tree2