New plan for ProteinChemical issues

[gaurav]

At the 2026-04-02 DOGSLED meeting, we decided on the following plan:

• Come up with a list of mappings that would be applied if we did combine Proteins and Chemicals (into Proteins that contain chemical IDs, not Chemicals)
  • We have Protein/Chemical duplicate CURIE lists (An ID can be present in multiple cliques (CURIE duplication) #276), but generating these from scratch probably makes more sense.
  • This would probably be easier to implement as a post-output step (use the mappings to combing chemicals and proteins as needed), but we should also look into moving the protein code into the chemical branch to simplify comparisons/xrefs -- the main thing to do there would be to modify the type determination code to identify polypeptides and proteins.
• Circulate to committees and get feedback.