A community-curated GFF file for the Platynereis dumerilii genome. The idea is that each expert in the community can add, refine, or amend gene models and the corresponding metadata starting from the most current Ensembl Metazoa version of the genome.
A GFF file is, first and foremost, a text file; and annotation consists of changing what certain lines of this file say ("this gene should be named hox3 instead of hox5"), or, in more extreme cases, adding or removing lines ("there exists a gene between positions A and B on this chromosome", or "this predicted gene is an algorithmic artifact and should be deleted"). To curate this as a community, we need a way to synchronously edit this file, discuss the proposed changes, perhaps vote on them, and, very importantly, keep track of them.
Git is a version control system that is very good at keeping track of changes to text files. GitHub is a platform designed specifically so people can collaborate on editing text files online. This is a match made in heaven.
Prerequisites: a GitHub account, some knowledge of the GFF format, love for Platynereis and/or phylogenies.
Contributing essentially consists of finding a gene in the GFF file and appending a proposed name as well as annotator information. You can follow along the example.
Every once in a while (depending on NCBI/Ensembl Metazoa update cycles) a new GFF version will be uploaded. This will trigger a minor version bump.
Sweeping changes to the assembly (like re-assemblies or similar) will be communicated via appropriate channels (spiraliabase, Platy community meetings, lab emails?).
The latest release will contain the consensus up-to-date version of the genome annotation. If this doesn't include your favorite gene feel free to use the version specified by the commit of the corresponding pull request. To avoid misunderstandings in publications, always refer to the genome by full version string and commit ID.
For technical reasons, the GFF file is split up into multiple files - one for the first 30 scaffolds, one for the rest of the scaffolds, and one for the mitochondrial genome. To create a full GFF file, simply concatenate these files. In a UNIX-based system:
In a UNIX system (Linux, MacOS):
$ cd /path/to/pdum-genome
$ cd chromosomes
$ cat *.gff3 > pdum-genome.gff3To create a GFF without the mitochondrial genome:
$ cat scaffold*.gff3 > pdum-genome-no-mt.gff3
$ cat unplaced.gff3 >> pdum-genome-no-mt.gff3In a Windows system:
(I did not test this, please let me know if it works/doesn't!)
$ chdir path\to\pdum-genome
$ chdir chromosomes
$ type *.gff3 pdum-genome.gff3To create a GFF without the mitochondrial genome:
$ type scaffold*.gff3 pdum-genome-no-mt.gff3
$ type unplaced.gff3 >> pdum-genome-no-mt.gff3Recent changes to the genome/repository are documented in the changelog.
- Find official GTF/GFF.
- Add a DOI to the repository.
- Tutorial on how to make changes.
- Automate version bumping.
- Automate DOI generation upon new release (zenodo?).
- Link to other resources:
- SpiraliaBase
- Ensembl Metazoa
- NCBI
- platynereis.com