GTDB-Tk
GTDB-Tk is a software toolkit for assigning objective taxonomic classifications to bacterial and archaeal genomes based
on the Genome Database Taxonomy (GTDB). It is designed to work with recent advances that
allow hundreds or thousands of metagenome-assembled genomes (MAGs) to be obtained directly from environmental samples.
It can also be applied to isolate and single-cell genomes. The GTDB-Tk is open source and released under the
GNU General Public License (Version 3).
Notifications about GTDB-Tk releases will be available through the GTDB Bluesky
account and the GTDB Announcements Forum.
Please post questions and issues related to GTDB-Tk on the Issues section of the GitHub repository. Questions
related to the GTDB can be posted on the GTDB Forum
or sent to the GTDB team.
🚀 Getting started
Be sure to check the hardware requirements, then choose your preferred method:
📖 Documentation
Documentation for GTDB-Tk can be found here.
🔄 What’s changed in v2.8.0
- No new reference data required: v2.8.0 uses the same R232 reference package as v2.7.x.
classify and classify_wf now report the same metadata. other_related_references lists only references with AF ≥ --min_af, closest first, and genomes not assigned by the ANI screen now report their closest references and warnings in classify_wf too (#717).- Classifications are unchanged except for fuzzy species in condensed genera, and only in
classify_wf (4 of 10,000 GlobDB genomes in our tests). In condensed genera, species are circumscribed more narrowly: their ANI radius is above the default 95%. If a genome falls outside the radius of its closest reference (AF ≥ --min_af), v2.7.x classify_wf could still assign it to a more distant species whose radius it falls within. v2.8.0 no longer does this; it applies the rule classify has always used.
- Example: a genome has 97% ANI to representative A, a fuzzy species with a radius of 98%, and 96% ANI to representative B (radius 95%), both with AF ≥ 0.5. A is the closest reference, but the genome is outside A’s radius. v2.7.x
classify_wf assigned the genome to B’s species; v2.8.0 (like classify) leaves the species unassigned (s__), lists A and B in other_related_references, and adds the warning that the genome falls outside the closest representative’s ANI radius.
- Output directories whose ANI screen was produced by an earlier version are re-screened when resumed.
See the change log for the full list.
✨ New Features
GTDB-Tk v2.7.0+ includes the following new features:
- Pre-sketched skani database: GTDB-Tk now uses a skani pre-sketched database of the GTDB representative genomes. This significantly reduces the database storage footprint from 198 GB (in Release 232) down to 98 GB.
- Representative genomes availability: The GTDB representative genomes are now available via the “Download” page on the GTDB website.
- Deprecated flag: Because the database is already sketched natively, the
--skani_sketch_dir flag is now deprecated.
- Replaced
--skip_ani_screen with --place_species: The --skip_ani_screen flag is now deprecated in v2.7.0 and has been replaced by the --place_species flag. The logic has been updated to reflect the new database structure:
- Previously: Using
--skip_ani_screen, genomes placed in a genus by pplacer were only compared to representative genomes within that specific genus.
- Now: Because the database is a single skani sketch, user genomes are compared against all GTDB reference genomes once at the very beginning of the pipeline. When the new
--place_species flag is selected, the genomes are still explicitly placed in the reference tree.
⚠️ IMPORTANT MEMORY WARNING: The divide-and-conquer approach now requires more than 128 GB of RAM. Specifically, you will need at least 140 GB of RAM for R232.
Using ANI screen “can” reduce computation by >50%, although it depends on the set of input genomes. A set of input genomes consisting primarily of new species will not benefit from ANI screen as much as a set of genomes that are largely assigned to GTDB species clusters. In the latter case, the ANI screen will reduce the number of genomes that need to be classified by pplacer which reduces computation time substantially (between 25% and 60% in our testing).
📚 References
GTDB-Tk is described in:
The Genome Taxonomy Database (GTDB) is described in:
- Parks, D.H., et al. (2021). GTDB: an ongoing census of bacterial and archaeal diversity through a phylogenetically consistent, rank normalized and complete genome-based taxonomy. Nucleic Acids Research, 50: D785–D794.
- Rinke, C, et al. (2021). A standardized archaeal taxonomy for the Genome Taxonomy Database. Nature Microbiology, 6: 946–959.
- Parks, D.H., et al. 2020. A complete domain-to-species taxonomy for Bacteria and Archaea. Nature Biotechnology, https://doi.org/10.1038/s41587-020-0501-8.
- Parks DH, et al. 2018. A standardized bacterial taxonomy based on genome phylogeny substantially revises the tree of life. Nature Biotechnology, http://dx.doi.org/10.1038/nbt.4229.
We strongly encourage you to cite the following 3rd party dependencies:
- Matsen FA, et al. 2010. pplacer: linear time maximum-likelihood and Bayesian phylogenetic placement of sequences onto a fixed reference tree. BMC Bioinformatics, 11:538.
- Shaw J. and Yu Y.W. 2023. Fast and robust metagenomic sequence comparison through sparse chaining with skani. Nature Methods, 20, pages1661–1665 (2023).
- Hyatt D, et al. 2010. Prodigal: prokaryotic gene recognition and translation initiation site identification. BMC Bioinformatics, 11:119. doi: 10.1186/1471-2105-11-119.
- Price MN, et al. 2010. FastTree 2 - Approximately Maximum-Likelihood Trees for Large Alignments. PLoS One, 5, e9490.
- Eddy SR. 2011. Accelerated profile HMM searches. PLOS Comp. Biol., 7:e1002195.
© Copyright
Copyright 2017 Pierre-Alain Chaumeil. See LICENSE for further details.
评论