run-LifemapR

Cassandra Bompard

2023-08-23

Introducing LifemapR, a package aiming to visualise data on a Lifemap basemap. The uniqueness of LifemapR is that it offers interactive visualization of data across a vast taxonomy (the NCBI taxonomy). LifemapR can easily process large datasets, exceeding 300,000 rows, with remarkable efficiency and precision. But what sets LifemapR apart is its flexibility and adaptability, offering a wide spectrum of customization options. This allows users to personalize their data visualization to better suit their unique needs and preferences.

1 Installation

  1. Install the development version of LifemapR from GitHub
remotes::install_github("damiendevienne/LifemapR")
  1. Once installed, the package can be loaded with :
library("LifemapR")

2 Usage

2.1 General case

The very first step to use this package is to have a dataframe containing at least a taxid column containing NCBI format TaxIds. This dataframe can also contain other data that you might want to visualise.

##      taxid                     X.Organism.Name BioProject.Accession
## 6  1004254         Cyberlindnera subsufficiens          PRJNA662200
## 7  1008252                                <NA>                 <NA>
## 8   100860                                <NA>                 <NA>
## 9   100903 Fusarium oxysporum f. sp. fragariae          PRJNA578477
## 10 1010633                    Digitaria exilis           PRJEB36539
##    BioProject.ID  Group
## 6         662200  Fungi
## 7             NA   <NA>
## 8             NA   <NA>
## 9         578477  Fungi
## 10        626237 Plants

You can then transform this dataframe into a format suitable for the visualisation functions of the package with the build_Lifemap function.

library("LifemapR")

LM <- build_Lifemap(data)

After that, you get a lifemap object that takes the form of a list containing the following informations :

You can then visualise your data with lifemap() associated with one or more of the following functions :

Each one of these three function adds a layer to the visualisation. These layers are combined with the + symbol.

# Example with default representation

# one layer
lifemap(LM) + lm_markers()

# three different layers 
lifemap(LM) + lm_markers() + lm_branches()

These function also allow the user to represent data by modifying characteristics of representations as we’ll see later in the examples.

The output is a shiny interface where the user can move and zoom freely.

Please note that the following examples have been done in may 2023, if you try them you may not have the same values due to database update

2.2 Analyse of Kraken2 results

This dataset is the result of a classification by Kraken (Derrick E. Wood, J. Lu, 2019) on a metagenomic sample coming from a controlled set of 12 known bacterial species (V. Sevim, J. Lee, 2019).

First of all, we load the data and transform it into the LifemapR format.

data(kraken_res)

LM_kraken <- build_Lifemap(df = kraken_res, basemap = "ncbi")

Then we can began to visualize our data.

We can for example represent the number of read that was assigned to each TaxID by the color of the markers with the following command.

lifemap(LM_kraken) + 
  lm_markers(var_fillColor = "coverage_percent", fillColor = "PiYG")
visualisation of kraken data

visualisation of kraken data

With this representation, the markers are displayed if the if the associated node is close enough.

It is possible to change the way markers are displayed with the display argument.

# All the nodes that were requested by the user
lifemap(LM_kraken) + 
  lm_markers(var_fillColor = "coverage_percent", fillColor = "PiYG", display = "requested")

# Only the nodes that have no descendants
lifemap(LM_kraken) + 
  lm_markers(var_fillColor = "coverage_percent", fillColor = "PiYG", display = "leaves")
left : display = "requested", right : display = "leaves"left : display = "requested", right : display = "leaves"

left : display = “requested”, right : display = “leaves”

Informations can also be displayed for the user with the popup or label arguments.

# When clicking on a node, display the desired information
lifemap(LM_kraken) + 
  lm_markers(var_fillColor = "coverage_percent", fillColor = "PiYG", popup = "name")
Usage of the ```popup``` argument

Usage of the popup argument

Finally, it is also possible to represent data with a subtree, either by the size of the branches or their color.

# Information on branche's color
lifemap(LM_kraken) + 
  lm_branches(var_color = "coverage_percent", color = "PiYG")

# Information on branche's size
lifemap(LM_kraken) + 
  lm_branches(size = "coverage_percent")