Note: This issue is tightly linked to https://code.google.com/p/phenotype-ontologies/issues/detail?id=51
The new uberpheno misses only 730 subclass links from the old version (F1000). To give
examples
- Short upper lip (HP:0000188) -> abnormal(ly) decreased size lip (ZP:0010226)
- Elevated tissue non-specific alkaline phosphatase (HP:0010679) -> abnormal(ly) quality
blood (ZP:0000094)
- Delayed cranial suture closure (HP:0000270) -> Abnormality of fontanelles (HP:0011328)
(of course out of the 730, some of them are correctly missing as they were wrong in
the old version)
Anyway, 75 % of the 730 missing links are concerning ZP classes. I have two questions/hypotheses
(it would be easier to skype about this, but for now I will stick with email):
1) Again for most of the links I looked at, the addition of the abnormal qualifier
solves the issue (see attachment). It would be great if the abnormal qualifiers/components
could be added in batch. Shouldn’t this be done for hp-edit.owl AND for mp-edit.owl
in sync? It would be fantastic to see the abnormal-thing in every logical definition.
2) In hp-edit.owl and mp-edit.owl we make use of “ 'has component' some abnormal”,
but in vertebrate.owl these become "qualifier some abnormal”. Why is that? This seems
inconsistent to me? Shouldn’t we just update zp.owl to the new abnormal-pattern and
have consistent usage of “ 'has component' some abnormal”?
Reported by dr.sebastian.koehler on 2014-08-15 08:02:26
Originally reported on Google Code with ID 52
Reported by
dr.sebastian.koehler
on 2014-08-15 08:02:26