the owning compartment of an EPN overlapping two compartments should be determined by rules given in the spec

[adrienrougny]

The owning compartment of an EPN overlapping two compartments is left to the tool: "An EPN cannot belong to more than one compartment. However, an EPN can be drawn over more than one compartment. In such cases, the decision on which is the owning compartment is deferred to the drawing tool or the author." (section 3.5.4 in L1V2.0)

This decision should not be deferred to tools but rather follow rules given in the spec.
Otherwise maps become ambiguous (two same drawings by two different tools should lead to the same model at the end, in terms of biology).

[draeger]

How to deal with a true cross-membrane protein that is embedded in the membrane of two different compartments? Should the best practice be to assign it to the periplasmic space (if appropriate)?