Receptors
The definitions and tree placements of the receptor terms need revisiting.
For example, epidermal growth factor receptor activity ; GO:0005006' is described as 'combining with EGF to initiate a change in cell activity', but it's also a child of the kinase terms. The kinase activity of the receptor is separate to the EGF-binding activity and thus creates TPV.
Perhaps we should remove the receptor terms from the kinase part of the tree, and leave this to concurrent annotations?
Perhaps we should remove the grouping terms:
- transmembrane receptor activity ; GO:0004888 (describes both function and location)
- transmembrane receptor protein kinase activity ; GO:0019199
- transmembrane receptor protein tyrosine kinase activity ; GO:0004714
The problem is that a receptor molecule has multiple 'activities'. We need to check how we're defining the receptor terms and how to capture these multiple activities- either just by concurrent annotations, or by trying to place the terms in GO under multiple parents (tricky without creating TPVs).
We need to separate the name of a gene product (EGFR) with the activities of the gene product.
Related SF items:
https://sourceforge.net/tracker/index.php?func=detail&aid=2684750&group_id=36855&atid=440764