Reprocess and annotate OpenNeuro with SNOMED assessment terms

[surchs]

Context

We now support derivatives. OpenNeuro has derivatives (https://github.com/OpenNeuroDerivatives/OpenNeuroDerivatives), so we should add them. We are also changing the vocabulary for assessment tools to SNOMED in #202 and we will need to reprocess the existing OpenNeuro annotations to make sure we can still discover them

Why

• to keep OpenNeuro annotations in sync with our developing data models
• to make sure users can come to query.neurobagel.org to search for derivatives (not possible elsewhere atm)
• to test out in practice our new process for automatically processing OpenNeuro datasets
• so we are ready for users to add these annotations and then have them be reflected in our query quickly

Outcomes

• All currently annotated OpenNeuro datasets have a data dictionary that uses SNOMED terms for assessment tools (where applicable)
• There is an automatic process that generates new .jsonld files for each OpenNeuro dataset for which we update the data dictionary
• This automatic process also includes the derivatives (not sure if automatically extracted or semi-manually)
• The new .jsonld files are reuploaded to our OpenNeuro graph

potential issue list (temporary)

These issues should be moved to sub-issues

• review current process for updating the OpenNeuro node (how do we re-generate all the .jsonld)
• make sure all workflows on OpenNeuro-JSONLD still run (especially the ones merging upstream and creating new repos)
• Make a list of all used cogatlas terms in OpenNeuro (from graph or from dictionaries) and see how many of them have an equivalent in SNOMED
• create a process to index and keep up to date the list of ON datasets with derivatives. also figure out what kinds of pipelines (and versions) they have been run with (might require email)
• update the annotations of all OpenNeuro datasets we currently have annotated
• re-generate the graph files for the OpenNeuro node
• Redeploy the OpenNeuro node with the up to date tools