Discuss possible rules for using 'is small molecule activator/inhibitor' relations in the conversion

[vanaukenk]

To discuss and possibly refine:
if an input to a reaction has a ChEBI id, create a node for the chemical and, based on the relation between the chemical and the reaction, create an 'is small molecule activator' or 'is small molecule inhibitor' relation between the chemical and the downstream activity.

[ukemi]

When this is asserted, we already do this:
http://noctua.geneontology.org/editor/graph/gomodel:R-HSA-5653890?

IIRC we want to extend this to things like receptors when they are activated by chemicals:
For example reactions R-HSA-400434 and R-HSA-416530 are successive reaction in which the fatty acid receptor binds fatty acids and then signals to downstream g-proteins. In particular, the next reaction is the GEF activity of the activated receptor. We want a rule that the receptor has the small molecule activator of the fatty-acids.

Possible rule: If a binding reaction is upstream of a receptor activity and the input of the binding reaction is a chemical and the receptor, create a has_small_molecule_activator relationship between the receptor activity and the receptor.

Questions:

1. Are there cases where the upstream binding reaction is an inhibitor in Reactome? I don't think so. The flow wouldn't be correct
2. The GPCR in this Reactome model is assigned a GEF activity. This is true but could be deepened to the receptor activity. Do all GPCRs act as GEFs? According to the def they do. Perhaps we need an ontology ticket.