search

cellannotation / cas-tools

Cell Annotation Schema Tools
1 stars 0 forks source link

AnnData to CAS should pull general CAS metadata (if possible) #56

Open dosumis opened 3 months ago

dosumis commented 3 months ago 
author_name (string):  # Not in `uns` - but is on Discover. Is it available from `census` - maybe on collection?
author_contact (string, format: email): # Not in `uns`- but is on Discover.  Is it available from `census` - maybe on collection?
title: # There is a title in uns - but it is not always informative.  The collection will have a more informative title.  Can we retrieve as metadata from census?

Note - title missing from CAS - need to revive - something got screwed up in edits to general fields!

dosumis commented 3 months ago

If any of these metadata elements are not retrievable from Census we should

  1. Request improvements to census to add this information
  2. Extend anndata_2_cas to allow this to be specified in input.
dosumis commented 3 months ago

Update - citation info from uns is potentially useful e.g.

'citation': 'Publication: https://doi.org/10.1038/s41586-024-07069-w Dataset Version: https://datasets.cellxgene.cziscience.com/56a5c0ac-3f44-41a2-b99d-3157900f0d19.h5ad curated and distributed by CZ CELLxGENE Discover in Collection: https://cellxgene.cziscience.com/collections/45d5d2c3-bc28-4814-aed6-0bb6f0e11c82',

(Plus we may be under obligation to cite properly - so seems reasonable to store this whole string, as well as process if needed for CAS-fields.)

dosumis commented 1 week ago

@ubyndr I believe this is fixed (although don't see linked PR). If so, can you link to the relevant PR & close?

ubyndr commented 1 week ago

I'm pulling these fields into the RDF graph in pandasaurus_cxg. I'll quickly update the anndata2cas.

ubyndr commented 1 day ago

Needs to be rolled back