Consider modelling genome reference files in analysis_protocol

[hannes-ucsc]

… instead of analysis_process as we will be doing in v2 MVP.

Alternatively, reference files could be modelled as regular input files to an analysis_process in links.json but that would be highly duplicative.

While it is necessary to point at the actual genome reference files used for an analysis, it should also be possible for metadata consumers to determine the type of reference genome used (species, version, etc) without inferring it from the reference file names. Consider adding a property to analysis_protocol. I'm sure there is a matching ontology.

[Still need to fill the stuff below out]

For which schema is a change/update being suggested?

e.g.: I would like to request an update to the cell_line.json schema.

What should the change/update be?

e.g.: I would like to add a new field - random_integer - to this schema to allow data contributors to assign a random integer to the cell line.

This update constitutes a <major/minor/patch> change to the schema(s) it affects.

What new field(s) need to be changed/added?

e.g.:

• Field name: random_integer
• Field description: A random integer.
• Field type: integer
• Required: no
• Examples: 42
• CV or enum: no

Why is the change requested?

e.g.: Data contributors need to be able to assign a random integer to a cell line in order to keep track of it.