What's staying in postgres?

For now, we are keeping several types of annotations in postgres:

• ncRNA
• uncloned loci
• IEP evidence code - select ones; will need to add these back to Protein2GO and then delete from tazendra
  • IEP evidence code annotations made to proteins will eventually be manually deleted from postgres and entered into protein2go other specific annotations that have unique issues
  • fox-1 annotations that have yDp13 in with/from - will need to work out with uniprot how to handle these
  • jnk-1 and sir-2.1 annotations that have a WBTransgene in the with/from column - will need to work out with uniprot how to handle these
  • Cre-vbh-1 annotations - will be added to uniprot manually and then deleted from postgres
  • gst-10 - annotations will be added to uniprot manually using a doi as reference id and then deleted from postgres
  • MTCE.26 - annotations will be added to uniprot manually - need to update gp2protein file, but also need to check on status of uniprot identifiers in xrefs.txt file for elegans mitochondrial proteins
• Genes for which there isn't a 1:1 correspondence between gene and protein in UniProt (e.g., actins (fixed by Swiss-Prot; needs WB update), histones, eef-1A.1, .2, etc.)
  • These genes will just have to be dealt with on a case-by-case basis.

Archiving, Deleting Old Annotations

1. We created one new annotation, for mir-80, in the OA before we deleted all of the GO annotations housed in Protein2GO. We did this so that we'd have a new pgid (14232) from which to iterate new pgids going forward. We did this because the OA assigns pgid's by looking at the highest value and then iterating +1. Since the annotations remaining in the OA likely did not include the highest pgid, we wanted to create a new highest pgid so that if we ever had to restore the archived annotations, there would be no pgid collisions.

Back to Gene Ontology