search

BiodiversityOntologies / bco

Biological Collections Ontology
Creative Commons Zero v1.0 Universal
21 stars 3 forks source link

bco current state #120

Open andrecastro0o opened 2 years ago

andrecastro0o commented 2 years ago

hi, just wanted to check what is the current state of BCO. Is it being maintained? Are there development plans for the future? cheers

ramonawalls commented 1 year ago

I guess the length of time it took me to answer provides some sense of how well BCO is maintained. It is definitely still used in a number of projects, and I do maintain it, just slowly. Pull requests for improvements or additions are welcome.

I plan to refactor BCO to work with the Ontology Development Kit over the coming months, which I hope will make it easier to update and contribute to.

tucotuco commented 1 year ago

@ramonawalls @timrobertson100 Hi Ramona. I have wondered whether BCO might be a reasonable destination for the outcomes of the Diversifying the GBIF Data Model Project. I know we talked about BCO being a proving ground for Darwin Core ideas, where Darwin Core lacked the semantics. One possible outcome of the Diversifying Project that is on the table is informing a possible TDWG ontology. I would like to explore the viability of these connections at some point.

cmungall commented 1 year ago

Would it be easier to manage if BCO were a pure application ontology? It seems there are only a handful of native non-obsolete classes

tucotuco commented 1 year ago

I don't know if it would be easier to manage or not, but my interest is on the application side to be sure.

dr-shorthair commented 1 year ago

I have long thought that there should be an opportunity to align BCO with W3C SOSA/SSN and PROV-O. i.e. port SOSA/PROV into the OBO environment.

@cmungall what is a "pure application ontology" ?

wdduncan commented 1 year ago

@dr-shorthair I think by 'application ontology' @cmungall means that all the terms imported from other OBO ontologies (i.e., it doesn't have any terms of its own). Doing this, would mean that most of the ontology work would be centered on managing imports, rather than maintaining terms.

In principle, this may work out quite well for BCO. However, I am concerned about what happens if a new term is needed. Ideally, the new term would be submitted to the appropriate ontology. However, some ontologies are quite slow in acting on new term requests. This could create a lot of frustration on the BCO side.

Also, (@dr-shorthair) do you have an idea of how to align with W3C SOSA/SSN?

cc @robgur

dr-shorthair commented 1 year ago

Also, (@dr-shorthair) do you have an idea of how to align with W3C SOSA/SSN?

The structures in BCO are largely compatible with SOSA/SSN (and also with PROV-O). I believe SOSA/SSN attempts a slightly more general solution, but that these patterns could be easily re-implemented in the OBO environment, which would help with some of the issue that some of the biomedical community is grappling with (especially TDWG).

I would not propose importing SOSA.

ramonawalls commented 1 year ago

@tucotuco Let's chat somewhere outside this issue about DwC data model. I read through the draft paper and would like to learn more.

@cmungall Regarding converting BCO to an application ontology, I am not opposed to the idea, but there are some terms in BCO that I think would not have an obvious home in existing ontology. It is worth analyzing, though, and moving out whatever terms we can.

@dr-shorthair regarding alignment with SOSA/SSN and PROV-O, I am still quite interested in this. There has been some work lately to solidify the observation model in OBO (see https://jamesaoverton.github.io/qqv/primer.html), which I support and hope to bring in to BCO. There are still some detail gaps and it does not cover specimen collection or surveys, but it is at least a start in unifying the approach to observations across OBO ontologies. I asked James to consider SOSA/SSN when he was working this model, and from what I remember of SOSA, it should not be too difficult to align. If I find time, I could publish a mapping file using SSSOM, but I need to update BCO first.

dr-shorthair commented 1 year ago

@ramonawalls the 'SSN Extensions' model was designed to help a bit more - see in particular the alignment with OBOE

On my TODO list is a 'collection of samples' model in parallel with this.

cmungall commented 8 months ago

Here are my suggestions for homes for BCO classes:

OBI

ENVO

ramonawalls commented 8 months ago

Thank you, @cmungall for taking a concrete step to move this along. As I get more and more busy with other responsibilities, the idea of transitioning BCO to an application ontology grows ever more appealing.

I have been hesitant about moving some of the BCO processing into OBI, but it does seem that OBI has expended its scope somewhat. If the OBI editors agree it makes sense to move terms there, we may be able to deprecate BCO completely.

@dr-shorthair have you done any work already on aligning SOSA/SSN with OBO ontologies outside of BCO?

@tucotuco are you open to aligning the TDWG work with better established OBO ontologies, rather than BCO? We could still maintain it as an application ontology, but I would need to know what the use cases are for such a goal. @robgur @daijiang For our current work with PPO, we could simply import terms from OBI and ENVO, without having to use BCO.

cmungall commented 8 months ago

I think it would be good to get more explicit commitment on scope in OBI. I know a lot of non biomedical people might be hesitant to use it given the name. I wish it was OSI or just OI. But combining alike terms in one ontology will be easier to manage

On Mon, Oct 30, 2023 at 5:35 PM Ramona Walls @.***> wrote:

Thank you, @cmungall https://github.com/cmungall for taking a concrete step to move this along. As I get more and more busy with other responsibilities, the idea of transitioning BCO to an application ontology grows ever more appealing.

I have been hesitant about moving some of the BCO processing into OBI, but it does seem that OBI has expended its scope somewhat. If the OBI editors agree it makes sense to move terms there, we may be able to deprecate BCO completely.

@dr-shorthair https://github.com/dr-shorthair have you done any work already on aligning SOSA/SSN with OBO ontologies outside of BCO?

@tucotuco https://github.com/tucotuco are you open to aligning the TDWG work with better established OBO ontologies, rather than BCO? We could still maintain it as an application ontology, but I would need to know what the use cases are for such a goal. @robgur https://github.com/robgur @daijiang https://github.com/daijiang For our current work with PPO, we could simply import terms from OBI and ENVO, without having to use BCO.

— Reply to this email directly, view it on GitHub https://github.com/BiodiversityOntologies/bco/issues/120#issuecomment-1786254897, or unsubscribe https://github.com/notifications/unsubscribe-auth/AAAMMOJVXCM5LCLIL5DG4ODYCBBUBAVCNFSM52ZIBPEKU5DIOJSWCZC7NNSXTN2JONZXKZKDN5WW2ZLOOQ5TCNZYGYZDKNBYHE3Q . You are receiving this because you were mentioned.Message ID: @.***>

dr-shorthair commented 8 months ago

It's been a while @ramonawalls .

Last time I tried, I found myself looking mostly at BFO and OBI, rather than BCO.

But I found it difficult because reviewing OBO properly requires setting up a full OBO environment etc to manage all the includes (I'm not routinely even routinely a Protege person). Then I got busy doing other things. Then I retired and went very part-time.

Meanwhile, O&M has been upgraded to OMS (Observations, Measurements and Samples) - see https://docs.ogc.org/as/20-082r4/20-082r4.html if you are prepared to deal with three layers of UML. And SOSA/SSN is currently in process of an update to match. Of particular note are

(i) three classes of sample: spatial-, material-, and statistical-sample (ii) ObservationCollection and SampleCollection, which should have some level of homogeneity, and may each be nested (heading a bit datacube) (iii) some alignment with IoT standards already in place in OGC - systems, platforms, deployments.

I also did a bit of work with GBIF, helping them move to an 'event based' model to disentangle processing and interpretation chains. This was also one of the main topics of a recent Dagstuhl meeting.

dr-shorthair commented 8 months ago

Looking at https://www.ebi.ac.uk/ols/ontologies/obi I would slot the SOSA classes in as follows

(i) sosa:Observation rdfs:subClassOf [obi:planned process](http://purl.obolibrary.org/obo/OBI_0000011) . whose result is an estimate of the value of a quality (i.e. information) of some entity [obi:assay](http://purl.obolibrary.org/obo/OBI_0000070) rdfs:subClassOf sosa:Observation .

(ii) sosa:Sampling rdfs:subClassOf [obi:planned process](http://purl.obolibrary.org/obo/OBI_0000011) . whose result is the creation of a Sample of some entity [obi:specimen collection process](http://purl.obolibrary.org/obo/OBI_0000659) rdfs:subClassOf sosa:Sampling .

(iii) sosa:Actuation rdfs:subClassOf [obi:planned process](http://purl.obolibrary.org/obo/OBI_0000011) . whose result is a change in a quality of some entity ([obi:animal feeding]() , [obi:freezing]() , [obi:material processing]() , [ogms:treatment](), _maybe some others_) rdfs:subClassOf sosa:Actuation

(iv) sosa:Sample rdfs:subClassOf [bfo:independent continuant] (http://purl.obolibrary.org/obo/BFO_0000004) . the result of a sosa:Sampling and having a isSampleOf relationship to another independent continuant.

ramonawalls commented 7 months ago

@jamesaoverton @bpeters42 I'm not sure who the most active OBI editors are at this time, but can you please take a look at the above thread and let me know if you think I should file an OBI ticket to move the discussion there?

bpeters42 commented 7 months ago

James and I are pretty good targets for active investigators. Based on the term list that @cmungall had on the 10/27 ticket, those would all be in OBI scope, and we would be more than happy to map them or take them on.

As an aside, I don't understand why the 'B' in OBI is a problem if it not a problem in in OBO (biology and biomedicine). Plenty of ontologies in OBO extend beyond that core purpose.

Let us know if you want to proceed, and then ideally submit a ticket to the OBI tracker.

ramonawalls commented 7 months ago

Thanks, @bpeters42 ! I'll create a ticket in the OBI tracker and tag people from this thread. Unless I hear otherwise, I'll create one overall ticket first for general discussion, then I can create tickets for individual terms.