R-script containing function to generate heatmap of pairwise dn/ds of value from dn/ds value calculated by MEGA program and other utilty functions for generating dn/ds heatmap with phylogenetic tree e.g function to generate list of correctly display scientific name (binomial nomenclature with italicized part for species name and non-italicized part for strain name) in plot.

This script was written for the article titled Evolution of organellar genes of chlorophyte algae: Relevance to phylogenetic inference.

MEGA is required for dn and ds value calculation. MEGA

This script depends on:

  ape version 5.2, ggdendro version 0.1-20, ggplot2 version 3.1.0, 
  grid version 3.4.4, phylogram version 2.1.0,  reshape2 version 1.4.3,  
  scales version 1.0, stringr version 1.3.1
  

Lower triangular matrix of dn or ds value from MEGA program can be read and converted into symmetric distance matrix using read_distancematrix function.

dn_sample <- read_distancematrix(dn_file_path)

Dendrogram in newick format can be read and converted into dendro type object in R with sorted leaf label using read_dendrogram function.

tree_sample <- read_dendrogram(tree_file_path)

The samples of input file can be found in sample_file folder.

MEGA program calculates dn and ds separately, so to generate dn/ds matrix we use dnds_matrix function. The order of label in matrix can be specified by specifying vector of ordered label to the dendrogram_label argument.

dnds_dataframe <- dnds_matrix(dn_mat=dn_sample, ds_mat=ds_sample, dendrogram_label=ordered_label)

To generate heatmap as shown in the article, we can use the dndsheatmap function. The function takes 4 arguments. dnds_mat: a symmetric matrix of dn/ds value taxon_list: the list of vector of species in each taxonomic group col_palette: the vector of colors that will be used to color the species name of each
taxonomic group; the order of color in vector are the same as the order of
taxonomic group in taxon_list argument ylabels: the vector of heatmap's y-axis labels

# taxon_list example 
taxon_list <- list('taxonomic_group1' = c('species1', 'species2', 'species3', 'species4'),
                   'taxonomic_group2' = c('species5', 'species6', 'species7'),
                   'taxonomic_group3' = c('species8', 'species9', 'species10')
                   )

# col_palette example
## red color corresponding to taxonomic group1
## green color corresponding to taxonomic group2
## blue color corresponding to taxonomic group3

col_palette <- c('red', 'green', 'blue') 

To plot dendrogram together with heatmap we use grid.newpage() function

grid.newpage()
pushViewport(viewport(layout = grid.layout(2,1)))
print(dendro_plot, vp = viewport(x = 0.585, y = 0.875, width = 0.524, height = 0.2))
print(heatmap_plot, vp = viewport(x = 0.5, y = 0.42, width = 0.8, height = 1.0))

We also provide the italic_sciname function to generate the list of scientific name that can be correctly display in proper text fomatting (binomial nomenclature with italicized part for species name and non-italicized part for strain name).

The function take the vector of name and the regular expression pattern to regcognize the italicized part and non-italicized part in scientific name.

# ([A-Za-z]+?\\s[A-Za-z.]+) to recognize scientific name
# (\\s[SAG|UTEX|UNA|CCMP|UTEX][A-Za-z0-9.\\s]*) to recognize strain name 
pattern <- '([A-Za-z]+?\\s[A-Za-z.]+)(\\s[SAG|UTEX|UNA|CCMP|UTEX][A-Za-z0-9.\\s]*)*'
name_vec <- c("Chlorotetraedron incus SAG43.81", "Neochloris aquatica UTEX B138")
ita_name <- italic_sciname(name_vec, pattern) 

Nuttapong Mekvipad [https://github.com/NMekvipad]