Modelling of unspecified regulation

[Andreas Gerasch]

Hi,

I have some problems with modelling unspecific activations and inhibitions of different physical entities using BioPAX L3, like Protein A activates Protein B.

I saw that some are using BiochemicalReactions for these regulations, but I think, this is not correct by definition. Another possibility is to have different states (modification features) for the entities (<none>, active, inactive, phosphorylated, etc.) and a Conversion between oppositional ones, the Conversion could then be controlled by a Control instance. The problem is, that whenever I try to model it according to the BioPAX standard L3, I introduce additional interpretation possibilities, which is not given by the source data.

Is there a common way for representing these unspecific regulations?

Thanks a lot and best regards,

Andreas

[Gary Bader]

Hi Andreas,
Activity flow relationships are not officially supported as they are ill defined, since the same entity can be an activator or an inactivator in different contexts. If you received this information from two different sources and wanted to merge it, you wouldn't be able to decide if two active molecules were the same physical entities. However, if this isn't important for your use case, you can encode this as A controls a conversion from B to B active.

Best,
Gary

http://baderlab.org
The Donnelly Centre - http://www.thedonnellycentre.utoronto.ca/
University of Toronto