taxit

taxit create, taxit check and taxit info from taxtastic (v1.0.1), the tool that builds pplacer’s reference packages, so that no taxtastic installation is needed. A refpkg comes out as taxit writes it: the same files, the same CONTENTS.json (contents, order, log, md5 sums) and the same phylo_model.json, parsed from the tree program’s log or stats file (RAxML, RAxML-NG, IQ-TREE, FastTree, PhyML). As with taxit, the creation date and the file name of phylo_model (a temporary file name) differ from run to run.

create: with a taxonomy, seq_info and tree, taxit reroots the tree with pplacer’s rppr reroot when rppr is installed and otherwise warns and skips it; gplace taxit create does the same (--no-reroot to skip it).
check: whether the refpkg is a valid pplacer input (its files, md5 sums, and that the alignment, Stockholm alignment, seq_info and tree name the same sequences), with taxit’s messages.
info: number of sequences and components, -n sequence names, -t tally per taxon, -l sequence lengths (taxit 1.0.1 crashes on -l; gPlace prints the table).

Arguments

usage: gplace taxit [-h] {create,check,info} ...

taxit subcommands

taxit_command: Possible choices: create, check, info

Sub-commands

create

Create a reference package.

gplace taxit create [-h] [-c] -P PATH -l LOCUS [-a NAME] [-d TEXT]
                    [-r VERSION] [-f FILE] [-i file] [-m FILE] [-M FILE]
                    [-p FILE] [-R FILE] [-s FILE] [-S FILE] [-t FILE]
                    [-T FILE] [--stats-type {PhyML,FastTree,RAxML}]
                    [--frequency-type {empirical,model}] [--no-reroot]
                    [--rppr RPPR]

Named Arguments

-c, --clobber: Delete an existing reference package.

Required arguments

-P, --package-name: Name of refpkg to create
-l, --locus: The locus described by the reference package

Package Metadata

-a, --author: Person who created the reference package
-d, --description: An arbitrary description field
-r, --package-version: Release version for the reference package

Input files

-f, --aln-fasta: Multiple alignment in fasta format
-i, --seq-info: CSV format file describing the aligned reference sequences, minimally containing the fields “seqname” and “tax_id”
-m, --mask: Text file containing a mask
-M, --model-file: File containing model information usually the .bestModel file
-p, --profile: Alignment profile
-R, --readme: README file describing the reference package
-s, --tree-stats: File containing tree statistics (for example RAxML_info.whatever”)
-S, --aln-sto: Multiple alignment in Stockholm format
-t, --tree-file: Phylogenetic tree in newick format
-T, --taxonomy: CSV format file defining the taxonomy. Fields include “tax_id”,”parent_id”,”rank”,”tax_name” followed by a column defining tax_id at each rank starting with root

Tree information

--stats-type

Possible choices: PhyML, FastTree, RAxML

stats file type [default: attempt to guess from file contents]

--frequency-type

Possible choices: empirical, model

Residue frequency type from the model. Required for PhyML Amino Acid alignments.

Taxonomic Rerooting

--no-reroot

Do not reroot the reference package using rppr reroot. [default: reroot if rppr is available and a taxonomy file is specified]

--rppr

Name of the rppr executable. [default: ‘rppr’]

Default: 'rppr'

check

Validate a reference package.

gplace taxit check [-h] REFPKG

Positional Arguments

REFPKG: Path to Refpkg to check

info

Show information about reference packages.

gplace taxit info [-h] [-n] [-t] [-l] refpkg

Positional Arguments

refpkg: the reference package to operate on

Named Arguments

-n, --seq-names: print a list of sequence names
-t, --tally: print a tally of sequences representing each taxon at rank RANK
-l, --lengths: print sequence lengths

Example

gplace taxit create -P my.refpkg -l bac120 -f msa.faa -t tree.nwk -s fasttree.log
gplace taxit check my.refpkg
gplace taxit info my.refpkg

Validation

Compared with taxit on 54 cases: refpkgs created from real logs of every tree program (19 files: FastTree JTT/WAG/LG/GTR, RAxML GTR/WAG/LG+F, RAxML-NG, IQ-TREE, PhyML DNA and protein with model or empirical frequencies), GTDB-Tk’s 8 class-level refpkgs rebuilt from their own files (their phylo_model identical to the one GTDB-Tk ships), every option (rerooting with and without rppr, --no-reroot, --clobber, --stats-type, errors), check on valid and 11 kinds of broken refpkgs, and info.