Location of receptor != location of its functions?

[dosumis]

See http://noctua.berkeleybop.org/editor/graph/gomodel:586fc17a00001124

Two issues:

1. Asserting the location of the function to be integral component of the membrane feels rather odd
2. The important functions occur on either side of the membrane.

The problem is even more striking for cell-adhesion molecules. In this case, if we assert that the binding between adhesion molecules on different cells occurs in the membrane, then we're actually asserting that the membranes of the two cells involved overlap. I prefer clean partonomies. These have already proven very useful for cleanly defining transport classes and sides of membranes.

@cmungall - With an all individual model LEGO now allows direct assertions of partonomy between gene products and CC. Should we encourage this in cases like this?

CC @thomaspd @ukemi

[ukemi]

Are retinioc acid receptors and other nuclear receptors also not being classified as receptors?

[dosumis]

Are retinoic acid receptors and other nuclear receptors also not being classified as receptors?

Seems to me this is a separate issue from the one on this ticket. An interesting point anyway though. Please see #42 .

[thomaspd]

We will make sure nuclear receptors are classified as receptors. See #38

[pgaudet]

@ukemi @vanaukenk
Did we agree that we will NOT be representing receptors like this ? ie - they are active at the plasma membrane (and not the extracellular part is active extracellularly, intracellular part is acting in the cyosol).

[ukemi]

This also came up in the Reactome work where we needed to make in inference about cases where a protein complex was located in a membrane and the reactants were, for example, located in the cytosol. We have terms like 'matrix side of inner mitochondrial membrane' that are part of the respective membranes. Using these terms allows us to specify the side of the membrane where the reaction occurs and associates the gene product with the membrane.