-
Notifications
You must be signed in to change notification settings - Fork 34
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Unstable output for owltools --save-closure-for-chado #256
Comments
Sorry, I got my file names swapped. I've edited the issue. |
This probably isn't very helpful, but I think it's something to do with "regulation of regulation " terms. |
Hi. In case it helps, we've found another example of this problem. The output of
regulates:
Sometimes we get this line in the output, sometimes not:
It changes when go-basic.obo changes, but not in a predictable way. The command line is:
The version of OWLTools seems to make a difference too. We've tried with the current master and v 0.3.0. The version build from the master branch doesn't give us the line above, v0.3.0 does. This was using the go-basic.obo with Is there any work-around for this problem that we could try? |
There's another example in this issue: geneontology/go-ontology#17171 |
Use -jar option to run from embedded jar. Possible fix for #256.
@cmungall @fbastian the behavior of this method is unstable (see above) across different Java versions (8 vs 11) and also sometimes jar packaging schemes:
Curious if you have any idea why that might be. I have verified that the same number of axioms are parsed when loading the file in all situations. |
I may have this figured out—I think it's to do with the usage of |
When I removed the |
We are working around this for PomBase by using a different tool to compute the closures: geneontology/go-ontology#17171 (comment) |
This is a follow on from pombase/pombase-chado#678
We're using
owltools --save-closure-for-chado
to generate the closure we load into Chado.@ValWood noticed that the output changes regularly for parts of the ontology that aren't changing.
We have been using these two files for testing because they are very similar:
https://curation.pombase.org/kmr44/go-basic-2018-05-07.obo
https://curation.pombase.org/kmr44/go-basic-2018-05-08.obo
We ran these commands:
Using owltools from the master branch, the output from
go-basic-2018-05-08.obogo-basic-2018-05-07.obo includes these lines:but the output from
go-basic-2018-05-07.obogo-basic-2018-05-08.obo doesn't have any lines that mention those two GO terms. We can't see any ontology changes that could cause this difference.With the owltools from May 2017 (version 447e415) that we have been using to load Canto, the output is quite different. For go-basic-2018-05-07.obo, the output is:
and for go-basic-2018-05-08.obo the output is:
In case it's useful here are the four output files:
https://curation.pombase.org/kmr44/owltools-chado-closure.tar.xz
(Also the recent version of owltools uses "RO:..." IDs for the relations but the older owltools uses term names like "positively_regulates". Is that a bug or should we upload our Chado loading code?)
@cmungall
The text was updated successfully, but these errors were encountered: