agree a controlled vocabulary for the section field

[only1chunts]

I believe v6.1 includes only 4 section names:

• sequencing ; terms that relate to how the DNA sample was processed for sequencing.
• environment ; terms that relate to the environment from which the sample/DNA was obtained.
• nucleic acid sequence source ; terms that are about where and how the DNA sample was created.
• investigation ; terms that relate to the project and linking it to other electronic resources.

Instead, I would like to propose the following updates:

• investigation - Terms that relate to the project or study and linking it to other electronic resources, these are not necessarily sample-specific terms.
• source (renamed from nucleic acid sequence source) - Terms describing the provenance of the nucleotide extract used in the experiment including any terms relating to archival sources of the physical specimen.
• sequencing - Terms that relate to the sample preparation and experimental processes leading to the generation of sequence data.
• environment - Terms relating to the descriptions of the surroundings and conditions from which the specimen was sampled
• methodology - Terms that relate to a method of data or metadata collection rather than any qualitative or quantitative values

[ramonawalls]

Currently in linkml, relation from term to subset is many to many? Should we make it one to one? I think that's how it was in the past. This does not preclude the term from being used in other sections when appropriate. We should explain this in the documentation.

[only1chunts]

1 - using the word subsetinstead of section is fine by me, it keeps things aligned with LinkML without changing any existing meaning to the way users may have thought about sections historically.

2 - @ramonawalls I'm not sure I follow your comment, you suggest it should be made 1:1 for term:subset, but then say that doesn't preclude a term from being used in multiple sections(subsets), surely thats exactly what 1:1 does preclude? I currently cant think of any examples of why a term should need to be part of multiple subsets (but am open to discussion), so I'm in favour of 1:1 mapping of terms to subsets/sections.

Added to CIG call agenda for discussion 25 June 2024.