I think the following basic cv terms are needed about mzqc HOT 4 CLOSED

Linked list can be tables, too, I don't mind - depends how we decide on the metric type issue. #73

from mzqc.

minor overlap with #71
IMO those could be added there.

from mzqc.

What we could formally recommend:

• Metrics values SHOULD have 'has_units' designation for each value type and SHOULD have 'has_type' ?
• Metrics MUST descend from a superfamily of with 'is_a'
• Can have other relations too, preferrably as 'part_of'
• Metric MUST have 'develops_from' ('derives_from') if it is intended as reimplementation
• Needed for a report can be described with cv term for a plot or report that 'has_part' of metric x, y, and z

from mzqc.

Do you want to add those metrics as general metrics? I.e. not linked to a specific tool or implementation?

For the recommendations, we'll need to explicitly specify how MUST and SHOULD should be interpreted.

• What does has_type encode? The type of metric, i.e. single value, list, etc., to replace the is_a way this is specified at the moment? Or something else?
• How do you define a "report", or "needed" even?

I'd caution against adding too much complexity to the schema, which seems to be happening here.

from mzqc.