ncbitax - a simple NCBI taxonomy parser

A very light-weight straight-forward parser for NCBI taxonomy, for simple lineage retrieval, and common ancestor searches.

Installation, Quick Start & Usage

With devtools from git

library(devtools)
install_github("raim/ncbitax")

Source code from github

• Download:

git clone git@github.com:raim/ncbitax.git

• Install:

R CMD build ncbitax
sudo R CMD INSTALL ncbitax_0.0.1.tar.gz

Use of this Package

This package allows to browse the NCBI taxonomy and extract sub-trees via sets of NCBI taxonomy IDs, eg. from a local blast database or genome-scanning results. The results can then be interpreted along the taxonomic tree or alignment-based phylogenetic trees, mapped to NCBI taxonomy IDs.

See the bash script scripts/setup.sh how to download NCBI taxonomy, database, the Silva 16S rRNA tree and a protein consensus tree from cyanobacteria and map all trees to NCBI taxonomy identifiers.

TODO

• improve speed by using more efficient indexed tree searches, eg. via separate package or Rcpp

• convert scripts/setup.sh to an Rmarkdown vignette for the R package, and move shell, Perl and R scripts to R package location for scripts

• generate a small sample taxonomy to allow examples to run

• getID: catch missing and resolve multiple hits

• add current NCBI taxonomy as data object or generate a subset, eg. cyanobacteria

• provide download & unpack function

• check whether functionalities are present in taxize package

• parse other formats, eg. newick, phylo object

• adapt and copy cmpTaxa.R and tax2newick.R rscripts from gIScanner to here

• add a high-level function: mapTaxa to find the shortest paths from query to target taxon IDs