dualingTrees R Package

James Meadow

5/19/2017

---

This package is designed to produce one type of visualization: phylogenetic effect sizes resulting from ecological interactions. This was designed to visualize mycorrhizal interactions, but could potentially be applied more broadly, as long as there are members of two phylogenies interacting in some way.

The work arose from a long-term MycoDB collaboration announced by Chaudhary et al. (2016; https://www.nature.com/articles/sdata201628). Additional publications are forthcoming, including publications surrounding the data used for examples within this package.

Installation:

Currently only housed on GitHub, so you can install using devtools:

library(devtools)
install_github("jfmeadow/dualingTrees-pkg", subdir="dualingTrees")

Workflow:

You will need at least 3 objects to generate a figure:

• x_tree: a phylo object (see the ape package) for the tree that goes on the top axis.
• y_tree: a phylo object (see the ape package) for the tree that goes on the left axis.
• And a 3-column data.frame consisting of:
  • x_key: names of organisms that match the tips of the x_tree.
  • y_key: names of organisms that match the tips of the y_tree.
  • response: the effect size or any other numeric interaction value to be displayed on the figure. These can be many replicate measures (to be averaged for each bubble size) or a single response value for each interaction combination (which would directly be plotted as a bubble).
  • Optionally, you can supply a categorical variable indicating the interaction type response_type. This will be used to color tree tips. Currently only 2 categories are allowed, with a third both category created on the fly when a single combination appears in two categories.

All other options are documented in the help files.

The package consists of 2 main functions that must be run in series, as well as a few internal ancillary functions. The typical workflow might look like this:

(Note: the example code will create a png into your current working directory)

library(dualingTrees)

## these are example datasets included in this package
data(fungal_tree_blup)
data(plant_tree_blup)
data(blup)

blup_trees_list <- 
  input_trees(
    x_tree = fungal_tree_blup,
    y_tree = plant_tree_blup,
    x_key = blup$fungus,
    y_key = blup$plant,
    response = blup$intrcpt,
    bubble_scale=5, 
    bubble_transform='sqrt', 
    y_lab_cutoff = 0.2,
    y_node_labs = c('fabaceae', 'eucalyptus', 'acacia', 'pinaceae'))
    
plot_trees(
  trees_list = blup_trees_list,
  pn_cols = c('#409ab1', '#dd1d18'),
  x_tree_col = 'gray40',
  x_bar_axis_offset = .6, 
  y_bar_axis_offset = .6, 
  x_space = .3, 
  y_space = .5, 
  png_filename = 'blup_trees.png', ## Figure in your working dir ##
  w_inches = 8, 
  h_inches = 8, 
  leg_text_pos = .6)

Which results in this figure (a png into your current working directory):

Here is another example, slightly more complex:

(Note: the example code will create a png into your current working directory)

data(fungal_tree)
data(plant_tree)
data(mycor_ei)

trees_list <- 
  input_trees(
    x_tree = fungal_tree,
    y_tree = plant_tree,
    x_key = mycor_ei$fungal_name,
    y_key = mycor_ei$plant_name,
    response = mycor_ei$ei,
    response_type = mycor_ei$mycorrhizae_type,
    bubble_scale=5, 
    x_lab_cutoff = NULL,
    y_lab_cutoff = 1.4,
    x_node_labs = NULL,   
    y_node_labs = c('fabaceae', 'betulaceae', 
                    'rosaceae', 'myrtaceae', 
                    'asteraceae', 'poaceae', 
                    'pinaceae'))
    
## vector for x tree edge colors.
x_tree_cols <- rep('gray40', length(trees_list$x_tree$edge.length))
x_tree_cols[1:65] <- '#b154a0'
x_tree_cols[66:94] <- '#51ad4f'

plot_trees(
  trees_list = trees_list,
  pn_cols = c('#409ab1', '#dd1d18'),
  x_tree_col = x_tree_cols,
  x_type_cols = c('#51ad4f', '#b154a0'),
  y_type_cols = c('#51ad4f', '#b154a0', '#5a1b1a'),
  x_bar_axis_offset = .5, 
  y_bar_axis_offset = .5, 
  png_filename = 'ei_tree.png', ## Figure in your working dir ##
  leg_text_pos = .4)
  

And the result:

---

A few recommendations for creating dualingTrees visualizations with this young package:

• This was built for a dataset that has ~300 y-axis tree tips and ~50 x-axis tree tips. Divergence from this will move around axes and text and whatnot. The plotting function has some built-in axis movement arguments. If these don't fix it, contact the author or submit an issue.
• Look very carefully when using a new dataset to make sure the result makes sense. If you find a bug or something that looks wrong, contact the author or submit an issue.
• Each function has built in output for the user to make sure the result matches expectations. Pay close attention to assignment of colors and whatnot. If something is not as expected, contact the author or submit an issue.
• The plotting function was designed with dense, publication-ready graphics in mind. Thus it is set up to export to a png or pdf for predictable sizing and easy zooming and exploration.