FAIR, versioned releases of result trees

[rvosa]

The pipeline currently produces its final output as a Newick formatted text file that sits in the results folder of the pipeline. A data publishing strategy needs to be devised so that these results are published FAIRly for reuse. The following considerations are in play:

• the input and output data are public, no restrictions on reuse are foreseen (e.g. CC0)
• the output data is in a standard format, e.g. as per the EDAM ontology
• the ideal solution would automate the publishing so this can be done for every BOLD release without too much hassle
• the ideal solution would allow for direct downloads so that downstream pipelines can ingest the tree automatically
• which community repository is most suitable is unclear:
  • TreeBASE cannot handle trees of the projected size
  • GBIF is moving towards hosting trees, but is not (yet?) seen as the standard location
  • Dryad is quite opaque and less amenable to automated downloads

The current strategy is to push to Zenodo. This is considered 'done' once a solution is produced that is FAIR-for-machines.