Thank you Keith, I have started to work with Protege and you are right the user interface is much more usable then OBOedit; therefore and for the lack of alternatives, I will validate this answer. However, while working with Protege, I noticed that the expressive power of different ontology languages (OWL vs OBO) has a larger impact than I though. I have the impression, that OWL forces a way of thinking on me I am very uncomfortable with, I think I am too much used to the UML way of thinking. Examples, please correct me if I am wrong:
- Classes can't have properties, only individuals can (correct?)
- So, there is no way to "inherit" properties via the class hierarchy(?), e.g. all LifeCycleEvents depend on an "age" and "temperature" property, but I couldn't define these properties for LifeCycleEvent. Do I have to define them for any individual event (there are about 25), or is there some kind of inverse logic?
- Only instances can be used in the annotation, sometimes 2 specific state are indistinguishable, so the super state should be used.
Example: I might say: This animal is either in "Larva 1" or "Larva2" stage (both subclass of "Larva"), so I might annotate as "Larva".
I will read a bit more....
Btw, here is a quickly drafted model in protege's graph visualizer image