OBOFoundry.github.io
OBOFoundry.github.io copied to clipboard
OBOFoundry
Decide on action for FMA entry in OBO
Currently hardcoded to be the ancient obo format version I made. Should be made the complete ontology now it is in OWL. obo-basic can be a subproduct
cc @trapperkeeper
Now FMA natively produces OWL (not valid OWL2 but will be soon) do we just point directly to it? Do we create a distinct OBO-compatible version? (e.g. OBO PURLs, standard annotation properties and object properties)?
Of course my vote, speaking for the FMA team, would be to use the FMA directly. It is confusing having multiple differing versions in play.
From @trapperkeeper
URL: http://sig.biostr.washington.edu/share/downloads/fma/release/latest/fma.owl tracker: https://bitbucket.org/uwsig/fma/issues
What should the status of the OBO purls be? Both class purls and ontology purls?
- Keep these active, and have them direct to an 'OBO translation' of FMA
- Obsolete them
- FMA obsoletes its existing purls and joins the OBO system
We are effectively doing 1 right now. With the new FMA being natively OWL we can update the translation to be a minimal one of rewriting IRIs, thus the OBO shadow will be semantically equivalent
2 may well break lots of things
3 is up to @trapperkeeper but I suspect the answer is no
See also https://github.com/OBOFoundry/purl.obolibrary.org/issues/63#issuecomment-161012932
Sorry for the late response. I brought this issue up with the FMA team but, as suspected, we are not prepared to obsolete our existing purls. First, some of our purls are already in use, as data annotations, referenced from other sources like UMLS, etc. Second, there are multiple community ontology efforts out there, and sometimes professional tension between them. The FMA, in an attempt to remain neutral, has adopted a purl that is not aligned with any specific effort (just trying to remain Switzerland here).
Of course, there are factors that may affect FMA purls. The future of purl.org being one that we are watching.
Regarding choices 1 or 2, from our perspective if would, of course, be nice if there were only one set of purls. But, for legacy reasons and to maintain exiting references I can certainly understand that some mapping or translation method may be necessary.
Thanks
Just to point out one thing: purl.obolibrary.org is now completely independent of purl.org. You should not use purl.org for anything as the future is, as you hinted, in doubt. purl.org could vanish tomorrow and the obolibrary ontologies would not be affected
On 14 Dec 2015, at 13:16, trapperkeeper wrote:
Sorry for the late response. I brought this issue up with the FMA team but, as suspected, we are not prepared to obsolete our existing purls. First, some of our purls are already in use, as data annotations, referenced from other sources like UMLS, etc. Second, there are multiple community ontology efforts out there, and sometimes professional tension between them. The FMA, in an attempt to remain neutral, has adopted a purl that is not aligned with any specific effort (just trying to remain Switzerland here).
Of course, there are factors that may affect FMA purls. The future of purl.org being one that we are watching.
Regarding choices 1 or 2, from our perspective if would, of course, be nice if there were only one set of purls. But, for legacy reasons and to maintain exiting references I can certainly understand that some mapping or translation method may be necessary.
Reply to this email directly or view it on GitHub: https://github.com/OBOFoundry/OBOFoundry.github.io/issues/21#issuecomment-164562226
Any purl scheme is a tie to a purl provider, be it purl.org or purl.obolibrary.org. We chose purl.org some time ago because we thought, at the time, that it would be the most universal and neutral provider. We expected the purl user community to embrace the unbranded purl and therefore expected it to be a stable resource. They are now on the ropes, as you say. However, we are already using them. Those FMA purls are already out there in the world. So, until we know its fate for sure, we are holding out to see what happens to purl.org. There is some talk of it being picked up and maintained by w3id.org: https://www.jiscmail.ac.uk/cgi-bin/webadmin?A2=ind1511&L=DC-ARCHITECTURE&F=&S=&P=3711 https://groups.google.com/forum/#!topic/persistenturls/Zpd4BHQxxIM
any more thoughts on this? http://purl.org/sig/ont/fma.owl still doesn't resolve
I can make it resolve to a web accessible copy of the OWL file itself, but it does not specify a version of the ontology.
OK, I have modified the PURL to resolve to the latest version (i.e. http://sig.biostr.washington.edu/share/downloads/fma/release/latest/fma.owl ).
I'm not sure of the context in which this arose, does this fit the bill?
-Todd
From: Chris Mungall [email protected] Reply-To: "OBOFoundry/OBOFoundry.github.io" [email protected] Date: Thursday, November 30, 2017 at 8:18 AM To: "OBOFoundry/OBOFoundry.github.io" [email protected] Cc: trapperkeeper [email protected], Mention [email protected] Subject: Re: [OBOFoundry/OBOFoundry.github.io] FMA (#21)
any more thoughts on this? http://purl.org/sig/ont/fma.owl still doesn't resolve
— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or mute the thread.
Not sure why this was closed, the issue is still not resolved. OBO is serving an old version of FMA
I'd really like this to be fixed. Need the latest FMA for a couple of projects currently. Can we have this resolve to the obolibrary PURL in addition to any PURL preferred by FMA?
I notice the owl files are now provided zipped. I think this can work OK with at least some infrastructure if they are names FMA.owl.zip, but they seem to be just FMA.zip.
$ curl -L -s http://purl.org/sig/ont/fma.owl
<!DOCTYPE HTML PUBLIC "-//IETF//DTD HTML 2.0//EN">
<html><head>
<title>404 Not Found</title>
</head><body>
<h1>Not Found</h1>
<p>The requested URL /share/downloads/fma/release/latest/fma.owl was not found on this server.</p>
</body></html>
$ curl -L http://sig.biostr.washington.edu/share/downloads/fma/release/latest/fma.owl
<!DOCTYPE HTML PUBLIC "-//IETF//DTD HTML 2.0//EN">
<html><head>
<title>404 Not Found</title>
</head><body>
<h1>Not Found</h1>
<p>The requested URL /share/downloads/fma/release/latest/fma.owl was not found on this server.</p>
</body></html>
Hey all, I will stick a static copy of the FMA.owl at that address, as an interim solution. I can't do it until tonight, but will do so.
I grabbed a few moments away from what I was working on and posted that file. The PURL should work now (at least until the next FMA release which isn't pending).
unfortunately not
On Mon, Jul 27, 2020 at 9:16 PM Nomi Harris [email protected] wrote:
Thanks, @trapperkeeper https://github.com/trapperkeeper!
Can we now close this issue?
— You are receiving this because you modified the open/close state. Reply to this email directly, view it on GitHub https://github.com/OBOFoundry/OBOFoundry.github.io/issues/21#issuecomment-664767005, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAAMMOOAK6EI6DPHU5RRIALR5ZGLBANCNFSM4BNANH7A .
matentzn@mbp:~/tmp_data $ curl -L -s http://purl.org/sig/ont/fma.owl | head
<?xml version="1.0"?>
<rdf:RDF xmlns="http://purl.org/sig/ont/fma.owl#"
xml:base="http://purl.org/sig/ont/fma.owl"
xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
xmlns:owl="http://www.w3.org/2002/07/owl#"
xmlns:xsd="http://www.w3.org/2001/XMLSchema#"
xmlns:rdfs="http://www.w3.org/2000/01/rdf-schema#"
xmlns:fma="http://purl.org/sig/ont/fma/"
xmlns:dc="http://purl.org/dc/elements/1.1/">
<owl:Ontology rdf:about="http://purl.org/sig/ont/fma.owl">
curl -L http://sig.biostr.washington.edu/share/downloads/fma/release/latest/fma.owl | head
% Total % Received % Xferd Average Speed Time Time Time Current
Dload Upload Total Spent Left Speed
0 0 0 0 0 0 0 0 --:--:-- --:--:-- --:--:-- 0<?xml version="1.0"?>
<rdf:RDF xmlns="http://purl.org/sig/ont/fma.owl#"
xml:base="http://purl.org/sig/ont/fma.owl"
xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
xmlns:owl="http://www.w3.org/2002/07/owl#"
xmlns:xsd="http://www.w3.org/2001/XMLSchema#"
xmlns:rdfs="http://www.w3.org/2000/01/rdf-schema#"
xmlns:fma="http://purl.org/sig/ont/fma/"
xmlns:dc="http://purl.org/dc/elements/1.1/">
<owl:Ontology rdf:about="http://purl.org/sig/ont/fma.owl">
These work now?
OK, PURLs work for me now
the question now is what to do with the class PURLs. We have OBO class PURLs for FMA from my translation that predates official FMA PURLs. The official FMA purls are not OBO formatted
Some options
- Deprecate OBO class PURLs, and "re-admit" FMA into OBO registry with the new PURLs as canonical
- Have shadow OBO class PURLs, as we do for NCIT
I am not sure how best to implement 1. I guess a one time release of the old OBO FMA version with all classes obsoleted and replaced by the sig PURLs. This would create an odd situation. the deprecated OBO class PURLs would be maintained within OBO infrastructure but the official FMA purls would exist without.
It's not really clear how meaningful it is to have FMA "in" OBO with 1.
I am not sure I am needed for any of this
I already made #1458 but no one commented on it :-).
I think we just need shared agreement between FMA and OBO
- does FMA exist outside OBO or within?
- if in, does FMA retain its current URIs? I assume yes. If so, then OBO will need to adjust its policy perhaps as a grandfather clause only. I think this will have to be discussed on an OBO call
- alternatively something the NCIT solution could be pursued but I suspect that will be deemed less than satisfactory
Do we need to arrange a call with someone at FMA to finalise and implement agreement on this? We need resolution for pending work with HCA and HubMap.
I think so but I would recommend HCA/HM just use the native FMA download PURL and class PURLs, and doesn't depend on OBO registry issues (we'd also need to coordinate with Uberon, and likely switch the Uberon bridge PURLs)
On Fri, Jul 16, 2021 at 2:20 AM David Osumi-Sutherland < @.***> wrote:
Do we need to arrange a call with someone at FMA to finalise and implement agreement on this? We need resolution for pending work with HCA and HubMap.
— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/OBOFoundry/OBOFoundry.github.io/issues/21#issuecomment-881304563, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAAMMOLOEFTFUPF4AQSEMHLTX72XHANCNFSM4BNANH7A .
What is the status of this? Does someone need to relay info to FMA? What about the linked draft PR, DO NOT MERGE: one solution to FMA problem #21 #1458?