Use cases for institute info

[arschat]

Tier 1 definition has an institute field described as

Institution where the samples were processed.

We currently map this to process.process_core.location that connects specimen_from_organism to cell_suspension, but it is not a easily accessible field. We've discussed to introduce a separate field for this.

Biomaterial level

We have to make an assumptions here:

1. Tier 1 samples might have some processing applied (i.e.cell_enrichment is applied in sample) but DCP specimen_from_organism is the unprocessed specimen before any protocol is applied (maybe preserved or stored at most).
   If we want to record the institute in the specimen/sample level, we have to assume:
   1. all process is done on the same institute
      (this might include dissociation, enrichment, cell_line differentiation, organoid generation, imaging etc.)
   2. different process institutes below cell_suspension (i.e. library_prep, sequencing, analysis) could not be represented if institute is recorded at the sample

This is rarely explicitly mentioned in publications, but most contributors might be able to provide this metadata.

Despite that the Tier 1 is not precise in terms of description, their intent is to bin this for batch effect per dataset rather than accurately describe the exact institute(s?) that processed the data. If this is the purpose each contributor should be able make the distinction of what differences in institutes processing samples, are important to state.

Type

In terms of type, it's most likely that a free-text with institute name would be sufficient. We can't really provide any controlled vocabulary for that here.

[arschat]

Record the institute in the cell suspension level in a free-text field.

[arschat]

PR in HumanCellAtlas/metadata-schema#1609