In GO annotation extensions, we should be able to enter gene IDs, and then have Canto and/or Chado loading magic convert to the optimal ID type (gene, transcript, or protein), depending on what's appropriate for the GO term.

This page is for drafts of a mapping that Kim can use to do the conversion. When Kim's ready to do so, and this mapping is more complete, move the content to a tab-delimited file in pombe-embl/chado_load_mappings/ .

For now, assume that the default will be the protein version of the ID unless specified otherwise in the mapping. So we won't need an entry for (e.g.) GO:0036211 protein modification process.

descendant of	ID type
GO:0003723 RNA binding	transcript
GO:0004540 ribonuclease activity	transcript
GO:0006355 regulation of transcription, DNA-templated	gene
GO:0016070 RNA metabolic process	transcript