search

obophenotype / zebrafish-phenotype-ontology

A vocabulary for describing Zebrafish Phenotypes
https://obophenotype.github.io/zebrafish-phenotype-ontology/
9 stars 3 forks source link

unsatisfiable classes when merged into vertebrate.owl #36

Open RLWOHIO opened 4 years ago

RLWOHIO commented 4 years ago

The latest release (2019-11-12) is a big improvement over the previous version. When standing alone, all classes are satisfiable. After being incorporated into vertebrate.owl though, some of them became unsatisfied, as listed below: ZFA_0000435 ZP_0000412 ZP_0002062 ZP_0002510 ZP_0002511 ZP_0002557 ZP_0003019 ZP_0003144 ZP_0003145 ZP_0003428 ZP_0004008 ZP_0004283 ZP_0004284 ZP_0004299 ZP_0005361 ZP_0005362 ZP_0005531 ZP_0005928 ZP_0006699 ZP_0006700 ZP_0006899 ZP_0006921 ZP_0006922 ZP_0008170 ZP_0008184 ZP_0008185 ZP_0008312 ZP_0008768 ZP_0009148 ZP_0010431 ZP_0010686 ZP_0010687 ZP_0010688 ZP_0011448 ZP_0011473 ZP_0011681 ZP_0011709 ZP_0011716 ZP_0011745 ZP_0011889 ZP_0012376 ZP_0012463 ZP_0012500 ZP_0013387 ZP_0014696 ZP_0014741 ZP_0014762 ZP_0014898 ZP_0015602 ZP_0016403 ZP_0016404 ZP_0016405 ZP_0016733 ZP_0016944 ZP_0016945 ZP_0016946 ZP_0017766 ZP_0018215 ZP_0018216 ZP_0018220 ZP_0018289 ZP_0018331 ZP_0019846 ZP_0019851 ZP_0019852 ZP_0019853 ZP_0019854 ZP_0020515 ZP_0020776 ZP_0020778 ZP_0021475 ZP_0021924 ZP_0022946 ZP_0100778 ZP_0100779 ZP_0100967 ZP_0102325 ZP_0102981 ZP_0102982 ZP_0103341 ZP_0103347 ZP_0103558 ZP_0103989 ZP_0106021 ZP_0107100 ZP_0107101 ZP_0108328

matentzn commented 4 years ago

Thanks! Some of the unsatisfiablities may only be solved much later in the year! Feel free to drop me an email so I can tell you about the general trajectory of uPheno and ZP -> we are working hard on pushing for a fresh, conflict free Unified Phenotype ontology by the end of the month, and I would love to get your input/requirements if you have any! :) vertebrate.owl will hopefully be deprecated before the summer.

RLWOHIO commented 4 years ago

This is great to know. Very much looking forward to having an error-free vertebrate.owl replacement. Since my work is primarily fish oriented, ZP and improved mapping of ZP terms to zebrafish genes are of great importance. I will continue to provide feedback as issues come up down the road. Thanks so much for your effort!

On Mon, Jan 6, 2020 at 7:23 AM Nico Matentzoglu notifications@github.com wrote:

Thanks! Some of the unsatisfiablities may only be solved much later in the year! Feel free to drop me an email so I can tell you about the general trajectory of uPheno and ZP -> we are working hard on pushing for a fresh, conflict free Unified Phenotype ontology by the end of the month, and I would love to get your input/requirements if you have any! :) vertebrate.owl will hopefully be deprecated before the summer.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/obophenotype/zebrafish-phenotype-ontology/issues/36?email_source=notifications&email_token=ADRGHRMIPIBJW4ZIPFAD5YLQ4MPFPA5CNFSM4J4Q4H5KYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEIFJRKA#issuecomment-571119784, or unsubscribe https://github.com/notifications/unsubscribe-auth/ADRGHRMYXZQ6EUDDNT54LMTQ4MPFPANCNFSM4J4Q4H5A .

matentzn commented 4 years ago

Can you describe a bit what you are using vertebrate.owl for that you cant get from simply using ZP? Which table to you use for the gene:phenotype annotations?

RLWOHIO commented 4 years ago

Vertebrate.owl serves as the foundation of semantic analysis of chemical toxicities. While my target species are fish, I have broad interest on other vertebrates, particularly human and mouse, and a range of biological domains as well. My last use (a year ago) of zf gene:phenotype associations came from https://github.com/obophenotype/zebrafish-phenotype-ontology-build. It seems to have been deprecated since.

On Thu, Jan 9, 2020 at 11:07 AM Nico Matentzoglu notifications@github.com wrote:

Can you describe a bit what you are using vertebrate.owl for that you cant get from simply using ZP? Which table to you use for the gene:phenotype annotations?

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/obophenotype/zebrafish-phenotype-ontology/issues/36?email_source=notifications&email_token=ADRGHRIY2YZBS7QVGRF2ZKDQ45DVBA5CNFSM4J4Q4H5KYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEIQ2GWY#issuecomment-572629851, or unsubscribe https://github.com/notifications/unsubscribe-auth/ADRGHRNLRWQQ5EGSZWX52KLQ45DVBANCNFSM4J4Q4H5A .

matentzn commented 4 years ago

So basically you are looking for a phenotype ontology that integrates all vertebrate phenotypes (and their anatomy, gene functions etc)? How exactly do you use vertrebrate.owl if you dont mind me asking? In a software tool? Or just to retrieve stuff using Protege?

EDIT: This is not official yet, but you can probably check

For the data you need now. Note, and this is very important, that these tables are under development and only work for abnormal phenotypes. There are some wrong associations in there between abnormal ZP classes and NORMAL phenotypes.

RLWOHIO commented 4 years ago

Yes. I use a java application I put together based on the OWAPI, reasoners, and the Semantic Measure Library ( https://www.semantic-measures-library.org/sml/). Thanks!

On Mon, Jan 13, 2020 at 7:31 AM Nico Matentzoglu notifications@github.com wrote:

So basically you are looking for a phenotype ontology that integrates all vertebrate phenotypes (and their anatomy, gene functions etc)? How exactly do you use vertrebrate.owl if you dont mind me asking? In a software tool? Or just to retrieve stuff using Protege?

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/obophenotype/zebrafish-phenotype-ontology/issues/36?email_source=notifications&email_token=ADRGHROOUYFCXSL2HE56YCDQ5RNJTA5CNFSM4J4Q4H5KYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEIYRCKI#issuecomment-573641001, or unsubscribe https://github.com/notifications/unsubscribe-auth/ADRGHRPEG5BMIYZX6ZFPPILQ5RNJTANCNFSM4J4Q4H5A .

matentzn commented 4 years ago

Can you describe your project in a bit more detail? Is your code on GitHub? What do you need from vertebrate.owl that you cant get from zp.owl all by itself?

RLWOHIO commented 4 years ago

My apologies for this delayed response. Our general interest is in the realm of mapping chemical-induced toxicity responses, particularly those of non-molecular nature, to their biology: genes, pathways, and diseases. This is why ontology-based semantic analysis is such an useful tool. While our focus is on fish and zp is great, having vertebrates included gives us a much wider coverage of biology and greater utilization of phenomics data available. The Java application is currently not available on Github. It is based on OWLAPI, Semantic Measure Library ( https://www.semantic-measures-library.org/sml/), and various reasoners, all publicly available.

On Thu, Feb 6, 2020 at 8:25 AM Nico Matentzoglu notifications@github.com wrote:

Can you describe your project in a bit more detail? Is your code on GitHub? What do you need from vertebrate.owl that you cant get from zp.owl all by itself?

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/obophenotype/zebrafish-phenotype-ontology/issues/36?email_source=notifications&email_token=ADRGHROD3YFXSDC2NU74RCDRBQFULA5CNFSM4J4Q4H5KYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEK7GTZQ#issuecomment-582904294, or unsubscribe https://github.com/notifications/unsubscribe-auth/ADRGHRKUMGCZ6VL72IIXW73RBQFULANCNFSM4J4Q4H5A .

matentzn commented 4 years ago

We are developing a new version of the Unified Phenotype Ontology - this will eventually replace vertebrate.owl and the rest of upheno1; I would be interested in your feedback and see whether it covers your use cases. Its not published yet, but a test version is available here. Please let me know what you think and what is missing from in there (or too much) to cover your case.

RLWOHIO commented 4 years ago

This is great. I would love to try it out when I get a chance. Thanks a lot for the offer.

On Wed, Mar 4, 2020 at 4:51 AM Nico Matentzoglu notifications@github.com wrote:

We are developing a new version of the Unified Phenotype Ontology - this will eventually replace vertebrate.owl and the rest of upheno1; I would be interested in your feedback and see whether it covers your use cases. Its not published yet, but a test version is available here https://ci.monarchinitiative.org/view/pipelines/job/upheno2/81/artifact/src/curation/upheno-release/all/upheno_all_with_relations.owl. Please let me know what you think and what is missing from in there (or too much) to cover your case.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/obophenotype/zebrafish-phenotype-ontology/issues/36?email_source=notifications&email_token=ADRGHRPLRMLZZZYQIOXERDDRFYQDHA5CNFSM4J4Q4H5KYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOENXDG4Q#issuecomment-594424690, or unsubscribe https://github.com/notifications/unsubscribe-auth/ADRGHRPCPCHOKDDZ2SJLAWTRFYQDHANCNFSM4J4Q4H5A .

RLWOHIO commented 4 years ago

I have done some testing.  On its own, upheno_all_with_relations.owl (03/04/2020), is fine.  Once merged with my annotations and relevant imports by OntoFox, there was a total 35 unsatisfiable classes, all from ZP.  This is a huge improvement over vertebrate.owl dated Dec 2019 which had a total of 455 unsatisfiable classes including 340 ZPs. Please see the attached for details.  Hope this helps. upheno_all_with_relations.owl.PLUS.mine2.owl.unsatisfiableZPs.pptx

matentzn commented 4 years ago

Ah very cool. So it seems that all the phenotypes that are unsatisfiable are so because of ZFA, which in turn interacts badly with UBERON. Can you delete the axiom:

image

And check whether that is the problem? If so, I will talk to RO and UBERON and we will decide on how to fix it.

RLWOHIO commented 4 years ago

Yes, you are right. By eliminating ZFA_0001284 and ZFA_0005491 from the merged ontology, all the other previously unsatisfiable ZP classes became satisfiable.

On Thu, Mar 12, 2020 at 1:20 PM Nico Matentzoglu notifications@github.com wrote:

Ah very cool. So it seems that all the phenotypes that are unsatisfiable are so because of ZFA, which in turn interacts badly with UBERON. Can you delete the axiom:

[image: image] https://user-images.githubusercontent.com/7070631/76547486-260c7500-6485-11ea-84e6-e6c731991670.png

And check whether that is the problem? If so, I will talk to RO and UBERON and we will decide on how to fix it.

— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/obophenotype/zebrafish-phenotype-ontology/issues/36#issuecomment-598311313, or unsubscribe https://github.com/notifications/unsubscribe-auth/ADRGHRMD5COFAAVUMJLUALDRHEKVTANCNFSM4J4Q4H5A .