Closed ghost closed 1 year ago
If we really need to refer to GEO IDs we should use https://bioregistry.io/registry/geo. => geo:GSE137537. Unless we can reliably tie the annotations to those on GEO, I don't think we should use GEO as a reference in cases like this. It would be better if we could use a standard CxG reference for the annotated matrix. It may be possible to use the UUID of the dataset, but we need some statement of official policy on IDs from CxG.
But in this case, we have DOI for all except one so I think we should use DOI alone for all except this and just drop the one with no DOI.
Context: #1974
The following edits are derived from review of the Retina bionetwork and cellXgene references: