-
Notifications
You must be signed in to change notification settings - Fork 21
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
Decide on format for MIxS URIs - namespace #233
Comments
We have discussed this multiple times on the MIxS as RDF working group calls. Below is a summary:
Therefore, we recommend using URIs of the format I will run this by the full CIG group at our next call. |
We still need to decide on the specific namespaces for terms, checklists, packages, and cvs. We discussed this at the hackathon in Vienna, but I can't find the notes. @jdeck88 or @pbuttigieg did you write it down? |
I don't think we want a separate namespace for each checklist. In the interest of simplicity, I suggest
Individual terms would then have IDs like: Packages would have URIs like Checklists would be similar https://w3id/gensc.org/mixs/human_gut.ttl`. |
Sounds good:
And to align the new IDs with the MIGS IDs,
see GOLD (Attached).
Cheers,
Lynn
…On Mon, Feb 10, 2020 at 2:25 PM Ramona Walls ***@***.***> wrote:
I don't think we want a separate namespace for each checklist. In the
interest of simplicity, I suggest
https://w3id/gensc.org/terms/ for individual terms
and
https://w3id/gensc.org/mixs/ for all checklist and packages
Individual terms would then have IDs like:
https://w3id/gensc.org/terms/MIXS_000001 .
Packages would have URIs like https://w3id/gensc.org/mixs/migo.ttl of
https://w3id/gensc.org/mixs/migo.xlsx.
Checklists would be similar https://w3id/gensc.org/mixs/human_gut.ttl`.
—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
<https://github.com/GenomicsStandardsConsortium/mixs-rdf/issues/2?email_source=notifications&email_token=ABBB4DMNXHMIEQCWF3WBVMDRCGSZBA5CNFSM4KGHYVG2YY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOELJ5AOY#issuecomment-584306747>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/ABBB4DNKYTOEP7VAPY6ITS3RCGSZBANCNFSM4KGHYVGQ>
.
--
Lynn M. Schriml, Ph.D.
Associate Professor
Institute for Genome Sciences
University of Maryland School of Medicine
Department of Epidemiology and Public Health
670 W. Baltimore St., HSFIII, Room 3061
Baltimore, MD 21201
P: 410-706-6776 | F: 410-706-6756
lschriml@som.umaryland.edu
|
We also need to consider controlled vocabularies (as values for attribute terms). Since the CV terms are still just terms, I suggest using the As a namespace for the lists of CVs, I suggest |
@pbuttigieg, does it make sense to have a namespace for MIGO that is separate from MIxS. In that case, MIxS packages and checklists would be in MIGO packages and checklists (is there more than one?) would be in and all terms (regardless of where they are used) would be in |
I'm not sure MIGO needs a separate namespace. |
Agreed.
…Sent from my iPhone
On Feb 28, 2020, at 3:22 PM, John Deck ***@***.***> wrote:
I'm not sure MIGO needs a separate namespace.
—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub, or unsubscribe.
|
Important! Use this one for terms, packages, and checklists! Also, see below for packages and checklist version IRIs. The actual format should not have MIxS packages and checklists would be in Decision on 10 May 2022: We will use numerical IDs for each package and checklists. See the numbers at LINK. LinkML also makes the products of all packages and checklists. Let's make new IDs for those by concatonating the checklist number first then the package number, separated by an underscore. (Old comment, no longer valid: MIGO packages and checklists (is there more than one?) would be in https://w3id.org/gensc/migo) note: We may need to reconstruct some of the names to be sure there are no white spaces. They should all have dashes or underscores. and all terms (regardless of where they are used) would be in Term URIs should use the unique 7 digit string and follow this format: @sujaypatil96 - for your reference |
I expect that genomic observatory packages would be quite different than MIxS packages. I think that is why we proposed this at GSC in Vienna. For now, we will start with registering MIxS, and we can add MIGO namespace later if the need arises. |
Important! Use this one for packages and checklists with versions. Package and checklist URI namespace CORE: https://w3id.org/gensc/mixs/ - resolves to current version https://w3id/gensc/mixs/vX/ - resolves specific versions Continue to use numerical versions. Can add minor versions if needed. Checklists: https://w3id.org/gensc/mixs/checklist_name - resolves to current version https://w3id.org/gensc/mixs/vX/checklist_name - resolves specific versions Use acronyms. Packages: https://w3id.org/gensc/mixs/package/package_name - resolves to current version https://w3id.org/gensc/mixs/package/vX/package_name/ - resolves specific versions Use opaque IDs for these |
Yes, that's metadata about an observatory, rather than a sequence, so it will be quite different
These look good: in the packages, we can mix MIxS and MIGO terms, or terms from other dedicated namespaces. |
On Sep. 14, we agreed that package/checklist IDs should be numerical, seven digit. |
I just discovered this today: http://rs.gbif.org/sandbox/extension/mixs_sample.xml It seems at one point there was a vocabulary in use? E.g
these don't resolve. Whenever we set up the new system we should figure a way to make these resolve |
Thank you Chris,
we know about these ids.
Dealing with them/resolving them is on our AI for the MIxS release.
Cheers,
Lynn
…On Thu, Nov 19, 2020 at 7:32 PM Chris Mungall ***@***.***> wrote:
I just discovered this today:
http://rs.gbif.org/sandbox/extension/mixs_sample.xml
It seems at one point there was a vocabulary in use? E.g
- classes like http://gensc.org/ns/mixs/terms/Sample
- properties like http://gensc.org/ns/mixs/terms/env_biome
these don't resolve. Whenever we set up the new system we should figure a
way to make these resolve
—
You are receiving this because you were mentioned.
Reply to this email directly, view it on GitHub
<https://github.com/GenomicsStandardsConsortium/mixs-rdf/issues/2#issuecomment-730723785>,
or unsubscribe
<https://github.com/notifications/unsubscribe-auth/ABBB4DNNGB2RO7ZE2ZUJAPTSQW2KHANCNFSM4KGHYVGQ>
.
--
Lynn M. Schriml, Ph.D.
Associate Professor
Institute for Genome Sciences
University of Maryland School of Medicine
Department of Epidemiology and Public Health
670 W. Baltimore St., HSFIII, Room 3061
Baltimore, MD 21201
P: 410-706-6776 | F: 410-706-6756
lschriml@som.umaryland.edu
|
Thanks Chris. And I have mentioned to the GBIF folks that they need to use
the new ID format. I have filing an issue about it on my todo list.
|
I just filed gbif/rs.gbif.org#51 and GenomicsStandardsConsortium/mixs-rdf#31. |
Should this not be https://w3id.org/gensc/ ?
|
Thanks, @mdoering -- good catch! They should indeed be https with w3id.org as the top level. I will edit the comment above. |
This problem is covered in issue GenomicsStandardsConsortium/mixs-rdf#31 |
@ramonawalls just checking in ... have you put more thought into when you can register the MIXS namespace. |
@wdduncan I forgot that I actually registered with w3id.org four months ago. However, I think the issue here is when will the IDs resolve. We have been waiting for the release of MIxS6, but more packages keep coming, and it keeps getting delayed. I'm going to start a new issue to discuss this topic. |
This comment was marked as off-topic.
This comment was marked as off-topic.
BackgroundThis comment is about optimizing a user’s experience of accessing information about MixS terms via the LinkML auto-generated documentation pages. Users shouldn’t encounter any broken links, and GSC should use namespaces that are clear and authoritative, but also provide flexibility for the future. Management of these namespaces will involve Decisions regarding the URLs for MIxS resources were made in 2020, before we started using LinkML seriously. Our knowledge and understanding have changed since then, and the documentation of the decisions is not completely clear. @turbomam and @ramonawalls met to try to clarify what resources a given URL should redirect to (what content should be associated with a given URL), and we make a new proposal below. We have two layers of URLS - the verbose ones assigned by GitHub Pages and the terser ones available from w3id. Additionally, the LinkML files declare prefixes and their expansions. The GitHub Pages URLs are generated automatically based on the name of our GitHub organization and MIxS repo. Ideally, the LinkML prefixes would expand to w3id URLs which would be the official identifiers for elements of MIxS. We had included a namespace for Minimum Information about a Genomic Observatory (MIGO) The MIGO prefix is not yet assigned to any checklists or terms and is not currently required. If we wish to use the https://w3id.org/gensc/mixs namespace, then those Feb. 2023 ProposalI propose a solution that supports having future namespaces for standards or projects outside MIxS (e.g., for MIGO for global observatories) but is also easy to maintain and understand and works well with GitHub Pages and LinkML: gensc/mixs namespaceThe root namespace for MIxS would be
Per issue #533, we will change the name of our Github organization to
should redirect to
The key difference from the current state (after PR #531) is that “gensc” would be added to the w3id redirect. This follows the w3id recommended practice of using “https://w3id.org/$org/$project/ and allows for the future prospect of including other GSC standards or projects, such as MIGO (minimum information about a genomic observatory). MIxS checklists and extensions would use the same prefix. On 10 May 2022 we decided that checklists and extensions and all of their combinations would have numerical IDs. These numerical IDs have been assigned in the incomplete schemasheets branch but not in the current release yet. In the schemasheets branch, IDs for combinations of checklists and extensions are constructed by concatenating the checklist number first then the extension number, separated by an underscore. Following this proposal, our LinkML yaml file should be edited to generate documentation pages for checklist, extensions, and combinations of the following format: MIGS bacteria checklist: Soil extension (aka environmental package): Combination of MIGS bacteria and soil extension: The shorter URLs generated from w3id for those three pages would be, respectively:
Our original proposal was that all terms, regardless of whether they are used in MIxS or some other (putative) standard should come within the w3id namespace Per comments above, we expect the term sets for other standards that may arise to be quite different from MIxS (e.g., they would describe observatories, not samples or sequences), and they could still reuse/import terms from the MIxS namespace. Therefore, I think it is a more practical solution to simply use the MIxS namespace for MIxS terms. Under this proposal, URLs for terms (i.e. slots) would follow the same format as URLs for packages and extensions (i.e. classes). For example, the w3id URL for geo_loc_name (curie MIXS:0000010) would be:
And the documentation generated by GitHub pages would be:
Note a change from an earlier decision, where terms used URLs of the format https://w3id.org/gensc/terms/MIXS_0000001. Under the new plan, PURLs would follow the format https://w3id.org/gensc/mixs/0000001. In other words, we would change ‘terms’ to ‘mixs’ and we would not include Example of an actual term page for geolocation name (geo_loc_name): https://genomicsstandardsconsortium.github.io/mixs/0000010/ or |
What needs to happen: In LinkML, we should use the following prefix expansion rule:
Autogenerated documentation will make URLs for all MIxS element of the form:
for any element of the standard. These are the URLs that are embedded into the github pages. In w3id.org, we need to register the prefix
Example: For geo_loc_name
This is a one way redirect, in that the short URLs are not directly in the github documentation. The URLs are listed as curies instead. However, since w3ids are shorter and redirectable, we can use these as the "official" URLs for MIxS elements in publications. |
We still need to decide how to implement this. I think it should be in our schemasheets, rather than continuing to hand edit https://github.com/GenomicsStandardsConsortium/mixs/blob/main/gsctools/mixs_converter.py. We will document that process in a different issue. |
Go to this comment for solution: #233 (comment)
Although JSON does not strictly require term urls, much of what people need to do with mixs does (e.g., use mixs in linked data, use mixs terms in ontologies).
We discusses this topic at the CIG hackathon in Vienna in May.
Options include:
Use obo foundry purls
Make gensc purls
Make gensc URLs that are not purls (e.g., gensc.org/ns)
Keep namespace for terms in TDWG
Comment from @cmungall: Also https://w3id.org/
Comment from @lschriml: The last time we discussed this at the board level, there was a lot of support for:
Make gensc purls
--> Has this been discussed further on the CIG calls ?
Copy of GenomicsStandardsConsortium/mixs-ng#3
The text was updated successfully, but these errors were encountered: