Hi everyone,
We are excited to announce the release of Greengenes2 2024.09!
The release files including QIIME 2 compatible artifacts can be accessed from our FTP. The Naive Bayes classifiers were constructed with QIIME 2 2024.5 with scikit-learn version 1.4.2.
The primary goal of this release was a taxonomy revision to reflect the phylum level changes in the International Code of Nomenclature of Prokaryotes.
A side effect of this release is the taxonomy harmonization moved a manual process to a nearly fully automated one, which will simplify accounting for future GTDB and LTP releases. Though the process has changed, we observe remarkable consistency in the decorated taxon labels, with many of the same labels falling on the exact same phylogenetic node between releases.
Finally, our website does not yet reflect the new release but will soon. The database which backs the website is being rebuilt.
Major changes
First, the taxonomy has been completely rewritten and re-based off of GTDB R220 and LTP 08 2023. Critically, this means the taxonomy now uses revised nomenclature (e.g., Firmicutes -> Bacillota).
Second, mitochondria and chloroplast are now explicitly represented in the Naive Bayes classifiers and the backbone taxonomy (e.g., 2024.09.backbone.tax.qza
). Annotation of mitochondria and chloroplast is not accessible from the phylogenetic taxonomy as these sequence records were placed rather than part of the topology update.
Third, we expanded the set of ASVs represented in the placement and include another million or so.
Important
The topology of this release is identical to the 2022.10 release. What this means is that additional whole genomes are not represented in the phylogenomic backbone, and additional full length 16S are not included in the topology update. We are working on both of these items in parallel for a future Greengenes2 release. Because the topology is identical, alpha and beta phylogenetic metrics computed on 2022.10 will be identical to those computed on 2024.09.
Have fun, and please don't hesitate to send questions!
- Greengenes2 team