gPlace

gPlace places genomes on the GTDB reference trees and gives exactly the same result as pplacer, the placement program used by GTDB-Tk, while needing far less memory.

Note

Work in progress. gPlace is still being tested. Before you use its results in published work, compare them with pplacer’s on some of your own genomes (compare), and please report problems on the issue tracker.

Why gPlace?

  • Same results as pplacer. Every placement, likelihood and branch length is identical to pplacer v1.1.alpha19, the version GTDB-Tk uses.

  • Much less memory. The largest GTDB-Tk class-level tree needs under 44 GiB (pplacer: about 104 GB), and the full GTDB bacterial tree fits in 64 GB (pplacer: about 615 GiB).

  • Faster. On the same 64-core server, gPlace placed 10,000 genomes on the full GTDB R226 bacterial tree in 14 minutes; pplacer took 17 hours.

  • Checks itself. gplace check_install confirms that gPlace computes exactly as pplacer does on your computer.

Quick start

# install (see Installing gPlace)
git clone https://github.com/cmc-aau/gPlace.git
cd gPlace
pip install .
gplace check_install

# place the genomes of a GTDB-Tk alignment on a GTDB-Tk reference package
gplace place --msa_file user_msa_file.fasta --refpkg gtdbtk.package.2.refpkg \
             --out_file gplace.jplace --cpus 40

The Quick start page shows the other common tasks: the full GTDB tree, many alignments at once, and comparing with pplacer.

Citing

gPlace reproduces pplacer’s placements on the GTDB reference trees. If you use gPlace, please cite pplacer and GTDB (see References).

gPlace is open source, released under the GNU General Public License (Version 3). Questions and problems: the issue tracker.

Getting started

Using gPlace

Advanced

About