Process to change Crop for Compositional in CDNO

[LilyAndres]

Hi OBO team.

I have some questions.

We are discussing an issue in CDNO #51 @ddooley, @CropStoreDb, @kaiiam. It has been highlighted the need from different institutions and FoodOn members to use CDNO dietary nutritional component concentration terms for other 'material entities' and not only for crops or plants. For this reason, we are planning to change the pre-composed terms from the class 'nutritional component concentration in plant structure' for 'nutritional component concentration in organism' so maybe changing 'plant structure' [PO:0009011] for 'organism' OBI:0100026.

With the changes described above maybe could be good to change the name of the ontology from Crop DNO to Compositional DNO.

Would this be possible? And if so, the change from Crop to Compositional could be done by just editing the md file? Or do we need to follow a different process?

Thanks a lot.

[matentzn]

Changing the name is not a problem, yes, you can make a pull request in the MD file and that's it. I am more concerned about the scope. If you look at classes like http://www.ontobee.org/ontology/CDNO?iri=http://purl.obolibrary.org/obo/CDNO_0200010 and imagine then to apply to whole organisms, they are simply organismal traits.

The reason why this was not scope overlap before when you applied was that you said crop - indeed there is nothings specific enough for crops. But what about organismal traits in general? Here we have so much - OBA, VT, even phenotype ontologies. There is already so much chaos in this area, and we urgently need to reconcile it (which is already costly). Instead of broadening the scope of CDNO, would you consider fleshing out OBA instead? This will make your work immediately useful to other users of OBA! I can give you editors access and so on - you have already shown you can do your work well to trust you with it. What do you think?

[LilyAndres]

Thanks @matentzn, I think it makes sense. Before I make a decision, we need to discuss this with the team, especially I need to talk to @ddooley as we want to make sure we also provide terms for 'food materials' through the processing pipeline, so the nutritionist can also use these concentrations. I will back to you asap.

[matentzn]

Sorry Kai just told me on slack about my CDNO misconception. You are talking more about dead meat. In any case, maybe I was wrong and it's not really applicable- let's get some other opinions first :) thanks @LilyAndres

[kaiiam]

Thanks @matentzn I was just pointing out the intention here is to expand the nutrition scope of CDNO beyond plants to other types of foods.

I guess one could see concentration of glucose in a lobster either as from an organismal/physiological perspective or from a food science - I'm about to eat it - perspective. Here we're more concerned with the latter.

[Graham-J-King]

Just to add to the above, the base classes that underpin the generic aspect are 'concentration of' & 'dietary component'. These may be used to describe composition of materials throughout the food production (eg crops or organisms with legs or fins), harvest/cull, processing and supply chain (the 'farm' to fork concept). There is great interest and value in being able to trace and compare relative component concentrations with reference to a common vocabulary (which for subsets such as minerals extends to the growing matrix of soil/sea).

[jamesaoverton]

This is a good issue and a good discussion, but the right place for it is the OBOFoundry/OBOFoundry.github.io repository. This PURL repository doesn't actually track the information we're discussing, and not nearly as many people watch the issue tracker on this repo. So I'm going to use GitHub's "Transfer Issue" button to try to move the issue to the other repository.

[jamesaoverton]

That seemed to work! Sorry for the interruption, please continue.

[kaiiam]

Thanks @jamesaoverton

[LilyAndres]

Hey sorry we have been trying to discuss this issue; @ddooley, @CropStoreDb and I had a meeting last week. We agreed that within terms of CDNO 'concentration of dietary nutritional component' class, replacing 'plant structure' with BFO: 'material entity' is more appropriate and generic than replacing with OBI: 'organism'.

Thus:

'concentration of' [PATO:0000033] and inheres in some (
'dietary nutritional component' [CDNO:0000001] and
    part of some 'plant structure' [PO:0009011])

would become

'concentration of' [PATO:0000033] and inheres in some (
'dietary nutritional component' [CDNO:0000001] and
    part of some 'material_entity' [BFO:0000015])

Thoughts about it? @matentzn, @jamesaoverton, @kaiiam

Our justification is: CDNO 'concentration of dietary nutritional component' terms was originally intended primarily to be associated with datasets that quantify concentration of chemical nutritional components derived from plant samples through the crop production and food supply chain. However, feedback, since we established the draft and published set of component terms, is that these are equally valuable for a wider range of organismal (e.g. beefsteak) and non-organismal (e.g. table salt) food materials throughout the processing and supply chain. There is no other ontology in place that provides a comprehensive set of such terms with relationships that we have established relevant to the domains (e.g. production, supply, and nutritional food composition).

In terms of use of the generic CDNO compositional terms, as outlined in the original scope, this is intended to be associated with quantification data, but also functional attributes where terms would be reused from other domains.

More generally, quantification data are widely used and reported in research literature, to inform food composition tables and labeling, or for supply chain quality assurance and control. However, the collective value is not realised due to a lack of agreed vocabularies. There is a range of major initiatives that are aiming to survey and generate new data, with which we have consulted, and this is a universal issue.

[kaiiam]

@LilyAndres plan seems reasonable, but changing the axiom is a discussion internal to CDNO it doesn't need to be discussed here in the OBOFoundry github. Here we're simply asking if we can change the ontology name (Crop to Compositional) and expand the scope to nutritional terms of species beyond plants. @matentzn and @ddooley is this reasonable?

[LilyAndres]

Thanks @kaiiam, I thought we were discussing the matter:

The reason why this was not scope overlap before when you applied was that you said crop - indeed there is nothings specific enough for crops. But what about organismal traits in general? Here we have so much - OBA, VT, even phenotype ontologies.

Just wanted to make sure that if we decided to use 'material entity' it doesn't overlap as highlighted by @matentzn.

[Graham-J-King]

@kaiiam,@matentzn
As indicated by @LilyAndres and the justification therein, we are specifically requesting a change of CNDO name from 'crop' to 'compositional' and associated update in scope, in order to address requests to have this available for more generic use.

[kaiiam]

@CropStoreDb and @LilyAndres fair points I just meant we should overload this issue. I think you guys nicely summed it up in the last two comments.

@matentzn is this reasonable to proceed with?

[matentzn]

I personally can't really say anything, but I have reached out to the operations committee to discuss this; sorry I forgot to add it to the last meeting, hopefully this can be dealt with via email. Will let you know ASAP!

[matentzn]

Alright, there does not seem to be any official policy on changing one's ontology scopes. I think you have made your case, and the issue has been open for a while now, so I would say, go ahead with the renaming and see whether anyone screams. The steps are:

1. Make sure that the Markdown and (potentially) PURL configs have the correct information (i.e. make pull requests on these two files to change the names, and, potentially, the domains).
2. We leave those open for a week to see whether anyone screams, but I don't think so
3. Collaborate on potential conflicts in overlap in the future. I know you are probably aware of it, but I just want to re-iterate the principles of modularity (thanks @cmungall for bringing it up again). Many of the interoperability problems in the OBO-sphere are caused by poor modularity, e.g. cyclical dependencies between ontologies, extracting of terms from ontologies the term doesn't belong in the first place. I don't think this case here gives rise to such modularity issues, but it is good to be aware of it to understand our stake in this: being the guardians of modularity :)!

[alanruttenberg]

Curious, when you say "change the pre-composed terms from the class" do you mean keeping the IRI, or deprecating the old one and creating a new term?

[kaiiam]

Of course it's usually best to depreciate and replace, in this case they are still very new (only been up for a couple months) and a work in progress we might be able to keep the IRIs, but if that's not advisable then we'd just replace them.

On Tue, Apr 27, 2021 at 3:05 PM Alan Ruttenberg @.***> wrote:

External Email

Curious, when you say "change the pre-composed terms from the class" do you mean keeping the IRI, or deprecating the old one and creating a new term?

— 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/1483#issuecomment-827590697, or unsubscribe https://github.com/notifications/unsubscribe-auth/AC5QDSGKV3U6VZOC54E4CCDTK2ZAPANCNFSM427N4HLQ .

[alanruttenberg]

The way I would handle it is that If there's been an official release with the older terms I would deprecate and create new. If it's just been in an unreleased development version then there's no need. If it has been released and you are sure that you know all users of the ontology and have confirmed that they can accept the change, then changing without deprecating would be fine, but in that case I would update the release to have the changes as well so that other's don't pick it up and have the wrong definition.

[kaiiam]

Thanks @alanruttenberg that sounds good will do.

[LilyAndres]

Great! Thank you @matentzn much appreciated and thanks @alanruttenberg we’ll proceed as suggested.

[matentzn]

There is nothing more to be done here from the OBO side, so closing. Please feel free to reopen if there are questions.