Re-generation of imports leads to build errors

[drseb]

Hi,

with the re-run of make imports/xxx.owl, we are now having issues with logic inconsistencies. See e.g. https://travis-ci.org/obophenotype/human-phenotype-ontology/jobs/232460614 or http://build.berkeleybop.org/job/build-hp-edit/1946/console

This comes from one of the re-mireoted ontologies, i.e. chebi or uberon or GO.

Any ideas how to quickly fix this @cmungall ?

[cmungall]

The most common cause of this is two separate uberon imports in the chain, each coming from different versions. Need a quicker way to analyze these than looking in Protege..

@nicolevasilevsky can you take a look and generate an explanation, and see if the axioms are coming from 2 places?

[drseb]

@cmungall Why are there disjointness axioms in uberon? I thought we remove them for the import purpose?

[cmungall]

They're there for detecting errors. I don't recall removing them in imports.

…

On Mon, May 15, 2017 at 10:19 Sebastian Köhler ***@***.***> wrote: [image: screen shot 2017-05-15 at 19 18 08] <https://cloud.githubusercontent.com/assets/5172897/26070022/521bc1a6-39a3-11e7-8ebd-50ba8abf3d18.png> @cmungall <https://github.com/cmungall> Why are there disjointness axioms in uberon? I thought we remove them for the import purpose? — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#229 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AADGORIf0OduedT6iPtd4XTCyLKSTaZ7ks5r6IkfgaJpZM4NbWMr> .

[cmungall]

Related ticket: obophenotype/uberon#1189

It looks like there is a very stale set of uberon axioms kicking around in the import chain, conj-sac hasn't been part of the eyeball since 2015:
obophenotype/uberon@914304d

[drseb]

fixed