VII. Tools for tree and placement results visualisation

The jplace format

Most phylogenetic placements tools output their results using the "Jplace" JSON specification (FA. Matsen, 2012).

A jplace file is contains basically 2 informations:

• a newick representation of the tree to which queries were placed (with added branch labels)
• a table-like json representation of the information associated to each placement

Over the past years many tools have been developed to open and visualize the content of jplace files, both in terms of summarizing the placements statistics and in visualizing the tree itself decorated with placement results.

Where to find jplace outputs

Jplace files generated by PEWO can be browsed via a fixed directory structure. For instance, using the PAC procedure, the following structure will be built:

[working_directory]
\_[tool]
  \_[pruning]
    \_[conditions_A]
      \_ pruning_conditions_tool.jplace

A concrete example, placement results associated to pruning 0, using PPLACER, and parameters read_length=300;ms=6;sb=3;mp=40 can be found following this path:

[working_directory]
\_PPLACER
  \_0
    \_ms6_sb3_mp40
      \_ 0_r300_ms6_sb3_mp40_pplacer.jplace

To visualise the corresponding result, we will use iTol (but many other tools could be used, see next title). Go to https://itol.embl.de/upload.cgi and upload the jplace file you selected. By toggling the button associated to the "Datasets" -> "Phylogenetic placement" box on the right of the screen you will visualise both the reference tree and the placement results on the tree.

Opening jplace files for results visualisation

Graphical interfaces

• Archeoteryx: one of the first tool supporting the jplace format, allows to display placements results on a tree view. (https://sites.google.com/site/cmzmasek/home/software/archaeopteryx)
• iTol: an online tool allowing to load jplace files and particulalry versatile in displaying the annotated tree in different configurations and with different metadata. (https://itol.embl.de)

Command-line tools

• GUPPY: one the first tool specifically designed for the post-analysis and manipulation of placement results. Allows to manipulate jplace files, extract desired information (tree, subsets of placements...), to compare different samples via placement-based statistics and compute the phylogenetic diversity represented by the placements. (https://matsen.github.io/pplacer/generated_rst/guppy.html)
• GAPPAS: a faster reimplementation of many GUPPY tools but also novel functionnalities and options of tree visualisation (https://github.com/lczech/gappa/wiki)
• Phyloshift: a metabarcoding-focused tool, using placements for species identification. Contains several commands to manipulation placement results. (https://phylosift.wordpress.com/)

R libraries

• ggtree: a versatile library allowing to load jplace results, plot the corresponding tree and placements and build basic statistics about the placements (https://guangchuangyu.github.io/ggtree-book/chapter-treeio.html#jplace-format).
• BoSSA: a library dedicated to the visualisation of placement results (https://cran.r-project.org/web/packages/BoSSA/vignettes/bossa-analysis.html)

Other libraries

• genesis: a very optimised C++ library dedicated to ML-based phylogenetic analyses with functions dedicated to phylogenetic placements visualisation and analysis (https://github.com/lczech/genesis)