Added template for BioSchemas Tool annotation

[egonw]

trafficstars

Uses BioSchemas Tool 0.3-DRAFT (21 November 2018), see http://bioschemas.org/devSpecs/Tool/.

I have tested the output with https://search.google.com/structured-data/testing-tool and example data (for the RMassBank package) can be found in this gist: https://gist.github.com/egonw/8347966585990579260f3e7e59d3954f

The BioSchemas JSON-LD is embedded in a <script> element, and ignored by everything except that for looking for scripts. BioSchemas is an ELIXIR Europe-supported schema.org extension (Google, Bing, and friend) for finding metadata.

The patch is simple and clean, but I would not mind seeing it run on a dev version of the website first, for testing the deployment too. Basically, the way to do this would be to open the https://search.google.com/structured-data/testing-tool (again) and pass the URL to the live HTML page with this embedded JSON-LS. The output should look like this:

The warning that Tool is not recognized is expected (in the BioSchemas world).

[egonw]

I also looked into BioSchemas annotation for tutorials (CreativeWork) and tested that with the BridgeDbR package, and the results of that is written up in this blog post: https://chem-bla-ics.blogspot.com/2019/04/bioschemas-creativework-annotation-in.html

[mtmorgan]

This looks really interesting; is there more complete information that can be extracted, e.g., from the VIEWS page? This puts whatever info is in the DESCRIPTION file (e.g., some packages have a BugReports url), including Author / Maintainer info, as well as some information we include on the build, e.g., last git commit ids & dates.

[egonw]

I think there is indeed more information the process generating this HTML (and the JSON-LD) has access to. I did not want to make it too long in the first instance (and restrict to info that was always there), but will write up some possible extensions this weekend. I may have to include some conditionals for that.

[egonw]

Sorry for the delay. I'm picking this up in the coming week.

[egonw]

Or so I planned... haven't lost track of it. I wanted to do that in the holiday I had then, but actually had a nasty cold/flu instead :(

[sneumann]

Hi, as mentioned above, this PR implements the basic, always available information into JSON-LD. Once merged, we'll see first uses of the Bioschemas stuff from package detail pages. As also mentioned above, there might be even more possibilities, but they'd require more hacking and logic. I suggest to merge now, and place further suggestions into an issue and hope for the next biohackathon. Yours, Steffen

[egonw]

Hi @mtmorgan, thanks for your patience. I finally got around to implement some of your suggestions and a few patches add a number of additional BioSchemas fields. There is more we can do, like link to Vignettes, some of that needing some extra function in https://github.com/Bioconductor/bioconductor.org/blob/master/lib/helpers.rb to output JSON-LD, easier in Ruby than in the HTML template.

But I prefer to roll out this first, and then do more advanced things later. Because with this out, I can get ELIXIR TeSS to start crawling and use this Bioschemas annotation in parallel with further development.

[lshep]

I apologize for the delay. This seems okay to merge. We would really like to have the maintainer field in the schema. Would it be possible if there is no maintainer in bioschema at least as a author. It should be available via @Package[:Maintainer] . once this is updated I will merge the request.

[egonw]

Oh, wonderful. I'll look at this asap! I have a couple of full day meetings this week, tho not sure how/what wrt to SARS-CoV-19, but hope to find time this weekend to add this. I don't think it was trivial, because it actually requires some rewriting of the structure. I will give an update soon.

[sneumann]

There are author and contributor (both cardinality MANY) in https://bioschemas.org/profiles/Tool/ but no explicit maintainer. Example below. Yours, Steffen

"author": {
    "@type": "Person",
    "familyName": "Kohlbacher",
    "givenName": "Oliver",
    "email": "[email protected]",
    "affiliation": "Wilhelm Schickard Institute for Computer Science, University of Tubingen"
  }

[mblue9]

This is great! pinging @abotzki who might also be interested to know about this. @aedin and I are interested in connecting Bioconductor with TeSS and I didn't know there had already been some work done on it.

@egonw I note this is from 4 years ago, are there any updates that need to be made?

@lshep could we try this out in the new website preview http://new.bioconductor.org (if that's kind of the dev site at the moment)

[lshep]

Not until the requested change was made. We are still waiting on the addition of author as requested above. The work stalled and there were no updates we were aware of

[egonw]

@egonw I note this is from 4 years ago, are there any updates that need to be made?

The pandemic kicked in (and this basically being a hobby didn't help much), and since September last year, we're back to "normal" but I'm still struggling to catching up with all the things that should have been completed during the pandemic.

But let's see where we are. Bioschemas itself is very much alive and I still (co-)maintain two Bioconductor packages.

Okay, I just checked and maintainer has been added, so, just like @sneumann's example above, we can have this now:

"maintainer": {
    "@type": "Person",
    "familyName": "Kohlbacher",
    "givenName": "Oliver",
    "email": "[email protected]",
    "affiliation": "Wilhelm Schickard Institute for Computer Science, University of Tubingen"
  }

[egonw]

@aedin and I are interested in connecting Bioconductor with TeSS and I didn't know there had already been some work done on it.

Yeah, this is great you like the idea too, like I did 4 years ago! I am if you take this forward and happy to act as tester, with my two packages.

Mind you, there are two kinds of annotation that we want:

• annotation of the tool itself (towards ELIXIR bio.tools, and the likes)
• annotation of the vignettes (and possibly other docs; towards ELIXIR TeSS)

To me, maintainer is more appropriate for use for describing the software package, while author can be used for both.

[sneumann]

Once you have an internet-reachable example page, the following could be used to check/validate: https://fair-checker.france-bioinformatique.fr/check Yours, Steffen

[mblue9]

Thanks for the insights and resources shared here. I'm looking into the structured data details for Bioconductor integration with TeSS and bio.tools. Also noted the EDAM discussion on https://github.com/bio-tools/biotoolsRegistry/issues/454. Will explore further as time allows. Appreciate all the input!

[abotzki]

Hi Steffen, thanks for the hint about the checker. It does not load at the moment. Who do we need to contact for this? Thanks.

On Wed, Nov 8, 2023, 10:25 Steffen Neumann @.***> wrote:

Once you have an internet-reachable example page, the following could be used to check/validate: https://fair-checker.france-bioinformatique.fr/check Yours, Steffen

— Reply to this email directly, view it on GitHub https://github.com/Bioconductor/bioconductor.org/pull/25#issuecomment-1801399247, or unsubscribe https://github.com/notifications/unsubscribe-auth/AJDCI5GDZBCMQYNXFCRQ4P3YDNFYTAVCNFSM4HDHVQYKU5DIOJSWCZC7NNSXTN2JONZXKZKDN5WW2ZLOOQ5TCOBQGEZTSOJSGQ3Q . You are receiving this because you were mentioned.Message ID: @.***>