BOOSTER - Help

Computing transfer supports via the web interface

Two types of jobs may be submitted to BOOSTER-WEB:

Complete phylogenetic analysis (datasets of moderate size);
Bootstrap support computation.

Complete phylogenetic analysis

If you do not have reference and bootstrap tree files, you can submit a multiple alignment to the run input form (Fasta, Phylip or Nexus format, may be gzipped with .gz extension only), in the “input sequence” field.

In that case, you can choose the workflow to run: (1) PhyML-SMS (for small/medium dataset); or (2) FastTree (for larger datasets).

These two workflows are installed and launched on the Institut Pasteur Galaxy server.

They constist of the following steps:

Tree inference, two possibilities: (a) Model selection + tree inference using Phyml-SMS; (b) Reference + Bootstrap Tree reconstructions using FastTree.
Bootstrap support computation using BOOSTER.

Bootstrap support computation

If you have reference and bootstrap tree files, you can also submit BOOSTER jobs directly (run page). In that case, two inputs are required:

A reference tree file in Newick format (may be gzipped);
A bootstrap tree file containing all the bootstrap trees (may be gzipped with .gz extension only).

Workflow

Please note that if a multiple alignment file is provided, no tree file will be taken into account.

Clicking the “Run” button will launch the selected analysis and redirect you to a results page, with the following steps:

The analysis will first be pending, waiting for available resources.
Then, as soon as the analysis is running, you will be redirected to a waiting page.
Once analysis is done, the results page shows the following panels:
1. Information about the run (identifier, start/end time, number of bootstrap trees analyzed, and output message).
2. Links to download the results:
  1. Tree with FBP (classical) supports.
  2. Tree with TBE (transfer bootstrap) normalized supports (download Newick format or upload to iTOL).
  3. Tree with branch labels formatted as follows: “Branch ID|Average transfer Distance|Size of the light side” (download Newick format or upload to iTOL).
  4. Booster log file with 2 parts:
    1. Instability score of every taxon (2 columns, “Taxon : Transfer Score”).
    2. Highly transferred taxa per branch (4 columns: Branch Id, Size of the light side, Average distance, and semicolon separated list of highly transferred taxa with their respective instability score).
3. Tree visualizer that highlights branches with a support (FBP or TBE) greater than the cutoff given by the slider.

Generating reference and bootstrap trees

If you want to generate reference and bootstrap trees via other means, you may do so using the following commands (example with 100 bootstrap replicates):

PhyML: Input file: alignment, Phylip format

phyml -i align.phy -d nt -b 100 -m GTR -f e -t e -c 6 -a e -s SPR -o tlr 
# Output Reference tree: align.phy_phyml_tree.txt
# Output Bootstrap trees: align.phy_phyml_boot_trees.txt

RAxML: standard bootstrap. Input file: alignment, Phylip format

# Infer reference tree
raxmlHPC -m GTRGAMMA -p $RANDOM -s align.phy -n REF
# Infer bootstrap trees
raxmlHPC -m GTRGAMMA -p $RANDOM -b $RANDOM -# 100 -s align.phy -n BOOT
# Output Reference tree: RAxML_bestTree.REF
# Output Bootstrap trees: RAxML_bootstrap.BOOT

RAxML: rapid bootstrap. Input file: alignment, Phylip format

# Infer reference tree + bootstrap trees
raxmlHPC -f a -m GTRGAMMA -c 4 -s align.phy -n align -T 4 -p $RANDOM -x $RANDOM -# 100
# Output Reference tree: RAxML_bestTree.align
# Output Bootstrap trees: RAxML_bootstrap.align

FastTree: you will need to generate bootstrap alignments (Phylip format), such as with goalign. Input file: alignment (Phylip or Fasta format)

# Build bootstrap alignments
goalign build seqboot -i align.phy -p -n 100 -o boot -S
# Infer reference tree
FastTree -nt -gtr align.phy > ref.nhx
# Infer bootstrap trees
cat boot*.ph | FastTree -nt -n 100 -gtr > boot.nhx
# Output Reference tree: ref.nhx
# Output Bootstrap trees: boot.nhx

IQ-TREE : booster supports and ultrafast bootstrap. Input file: alignment (Phylip format)

# Infer Reference tree + ultrafast bootstrap trees
iqtree-omp -wbt -s align.phy -m GTR -bb 100 -nt 5
# Output Reference tree: align.phy.treefile
# Output Bootstrap trees: align.phy.ufboot

Example dataset

You can try BOOSTER-WEB with the following trees inferred from the primate nucleotide alignment taken from “The Phylogenetic Handbook”:

Reference tree: .nw.gz
1,000 bootstrap trees: .nw.gz

If you also want to infer the reference and bootstrap trees:

Original alignment: .phy

After computingTBE and FBP using BOOSTER-WEB, you should obtain the following trees. TBE supports are larger than FBP supports for all branches. If bootstrap and reference trees were re-computed, some fluctuations of the supports are to be expected.

Installing a local version of the web interface

The web interface was developped in Go, and is thus executable on any platform (Linux, MacOS, and Windows). The only thing yo need to do is download the latest release of BOOSTER-WEB from Github, and run it by clicking on the executable.

Then visit the following url: http://localhost:8080.

Computing transfer supports via command line

BOOSTER is also available as a standalone executable (implemented in C). Source and binariy files are available on Github.

Usage: ./booster -i <ref tree file (newick)> -b <bootstrap tree file (newick)> [-d <dist_cutoff> -r <raw distance output tree file> -@ <cpus>  -S <stat file> -o <output tree> -v]
Options:
      -i : Input tree file
      -b : Bootstrap tree file (1 file containing all bootstrap trees)
      -a, --algo  : bootstrap algorithm, tbe (transfer bootstrap) or fbp (Felsenstein bootstrap) (default tbe)
      -o : Output file (optional), default : stdout
      -r, --out-raw : Output file (only with tbe, optional) with raw transfer distance as support values in the form of
                       id|avgdist|depth, default : none
      -@ : Number of threads (default 1)
      -S : Prints output logs in the given output file (average raw min transfer distance per branches, and average
      	   transfer index per taxa)
      -c, --count-per-branch : Prints individual taxa moves for each branches in the log file (only with -S and -a tbe)
      -d, --dist-cutoff: Distance cutoff to consider a branch for moving taxa computation (tbe only, default 0.3)
      -q, --quiet : Does not print progress messages during analysis
      -v : Prints version (optional)
      -h : Prints this help

Complete information regarding installation and usage is available on the GitHub page.

Data

The archive containing all data described in the paper can be downloaded from the github page.

Note

PhyML can also directly compute TBE supports (beta). To do this, you will need to download and install PhyML from its github repository:

phyml -i align.phy -d nt -b 100 --tbe -m GTR -f e -t e -c 6 -a e -s SPR -o tlr 
# Output tree with supports: align.phy_phyml_tree.txt

BOOSTER - BOOtstrap Support by TransfER