Skip to content
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.

Sign up for GitHub

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

Jump to bottom

Integrate 'organism' #6

Closed
jamesaoverton opened this issue Sep 21, 2018 · 36 comments
Closed

Integrate 'organism' #6

jamesaoverton opened this issue Sep 21, 2018 · 36 comments

Comments

@jamesaoverton
Copy link
Member

OBI:0100026 organism was never in OBI scope but was required for various tasks. It's defined as equivalent to Viruses or Bacteria or Archaea or Eukaryota.

I think NCBITaxon:131567 cellular organism is more recent. Its three subclasses are Bacteria, Archaea, and Eukaryota -- Viruses are not included.

I know that @bpeters42 has some thoughts on this.
@cmungall
Copy link
Contributor

Definition is a bit circular but I think this does the trick: http://purl.obolibrary.org/obo/CARO_0001010

Note you need one more axiom to make things link up:
obophenotype/caro#15

@bpeters42
Copy link
Contributor

bpeters42 commented Sep 21, 2018 via email

@zhengj2007
Copy link

+1 Bjoern

@cmungall
Copy link
Contributor

isn't OBI an odd place for this to live?

I'm 50-50 on name. On the one hand, we tend to err on the side of ugliness to prevent confusion. On the other, I especially don't like long union clauses. If the consensus is for short name then let's change it in CARO -- then would it make sense to import from CARO?

@cmungall
Copy link
Contributor

Note: need 'organism' to use in domain/range for biotic interaction relations in oborel/obo-relations#261

cmungall added a commit to oborel/obo-relations that referenced this issue Sep 26, 2018
@cmungall
Adding organism from CARO - may change label, see OBOFoundry/COB#6
e8c15a3
@bpeters42
Copy link
Contributor

bpeters42 commented Sep 26, 2018 via email

@cmungall
Copy link
Contributor

No criticism intended, just looking for the best long term home.

NCBITaxon is a good choice, except thus far we have treated this as a direct syntactic transform. See the docs on http://obofoundry.org/ontology/ncbitaxon.html

On the one hand I'd like to do some additional processing, e.g. depluralizing terms, but that would be a big break from our process so far (maybe it's time). We have also strictly adhered to the numeric part of the URI corresponding to a numeric identifier in the source ncbi taxonomy database. It gets a bit odd if we start minting our own URIs with the NCBITaxon prefix. First we have to choose a number that is guaranteed not to clash ('0'? Random 10 digit number?). Even then it's a recipe for sowing confusion...

@cmungall
Copy link
Contributor

We could just ask ncbi to add a taxon node for organism, they are our chosen authority here.

The foreseeable objection is that this may seen as a commitment about phylogenetic relationship of viruses in relationship to cellular organisms

@mcourtot
Copy link

I would strongly prefer not to mint new ID using someone else's prefix - if we really need new terms and NCBI taxonomy doesn't want to add them then the only compromise I can think of is adding them in the NCBITaxon file but with another prefix (something like OBOTaxon or else) so they'd be available but clearly not created/supported by NCBI.

I like the idea of asking NCBI to add an organism that includes cellular organisms + virus + viroid - should we try that and advise if they're reluctant? @cmungall do you know the right person to ask?

@nataled
Copy link

nataled commented Sep 26, 2018

Perhaps NCBITaxon:1 ("root") would somewhat fit for the ID? True, the ontology does include non-organism things outside the scope of what we're looking for (such as plasmids), but since that's already the case maybe we can let it slide? Or, if additional processing is going to happen anyway, such terms can be removed.

@zhengj2007
Copy link

@mcourtot "I like the idea of asking NCBI to add an organism that includes cellular organisms + virus + viroid".

I like this idea too. One more question, in this case, shall we deprecated OBI: organism?

@bpeters42
Copy link
Contributor

bpeters42 commented Sep 26, 2018 via email

@mcourtot
Copy link

Proposal is

  1. use the OBI term and change the logical definition to refer to cellular organism +virus + viroid. Add editor note on OBI term with details of the request in 2. We will keep 'organism' label for now.
  2. request new organism term from NCBI - when/if created advise about deprecating the OBI term

@cstoeckert
Copy link

cstoeckert commented Sep 27, 2018 via email

@zhengj2007
Copy link

+1 Melanie's proposal

@cmungall
Copy link
Contributor

Alternate proposal

  1. use the CARO class but relabel to 'organism'
  2. request new organism term from NCBI - when/if created advise about deprecating the CARO term

This just seems to make sense. organisms are anatomical entities, CARO is an upper-level anatomical entity ontology with root class 'anatomical entity'. OBI is an ontology of biomedical investigations, organisms exist independent of biomedical investigations. Full credit to OBI for providing the term since 2009 but we need to move forwards with a more sensible modularization strategy.

Another issue is that the OBI class explicitly makes no commitment to start/end, and explicitly says this is out of scope; the editor note says "10/21/09: This is a placeholder term, that should ideally be imported from the NCBI taxonomy, but the high level hierarchy there does not suit our needs (includes plasmids and 'other organisms'), 13-02-2009: OBI doesn't take position as to when an organism starts or ends being an organism - e.g. sperm, foetus. This issue is outside the scope of OBI." - for anatomy ontology use cases there needs to be ontological commitment.

@bpeters42
Copy link
Contributor

bpeters42 commented Sep 27, 2018 via email

@bpeters42
Copy link
Contributor

It looks like the next RO release is pushing for the CARO class to be the new 'organism'. As I said, I will no longer object to it. But then I would very much want to change that class to make it not less embarrassing

Currently:
label: "organism or virus or viroid".
definition: "Material anatomical entity that is a member of an individual species or is a viral or viroid particle."

Problems:

  • Label refers to 'organism or' without organism being defined anywhere. Minimally it should be "cellular organism or virus or viroid". Better yet: "generalized organism" Best: "organism"

  • Definition is inconsistent, which is just not apparent because there is a lack of logical constraints in CARO. 'material anatomical entity' is defined as being a kind of 'anatomical entity' which is part of a 'multicellular organism'. So clearly this class which covers non-multicellular organisms cannot be a kind of 'anatomical entity'. Parent should be 'material entity' entity instead.

  • reference to 'member of species' when species isn't defined buys us anything. And there species for viruses just as well as for cellular organism, so that doesn't help either.

Suggestion: Replace with the imperfect elucidation in OBI, dropping the 'living' part to stop the virus haters: "A material entity such as an animal, plant, bacteria or virus or viroid, that is capable of replicating or reproducing, growth and maintenance in the right environment."

@mellybelly
Copy link

Most definitions of organism do not include viruses and viroids. There is of course a whole body of literature arguing about how to define organism. I don't think we should spend a lot more time on this - an enormous effort was spent during the creation of these terms in CARO and other ontologies already and many decades more generally ;-).

I don't think this presence of this class/class label is a matter of being a virus hater. We do need a union class, and this is reason it was created, quite the opposite of being a virus hater.

I don't think 'organism' by itself as a label works as most organismal biologists use this term in exclusion of viruses and viroids and are very likely to use it incorrectly. I don't love 'generalized organism' but would be ok with that given that it is at least a flag that it might mean something more than simply 'organism'. I am also fine with keeping the union label.

Agree we should have definitions for organism, species.

Agree the text definition of 'anatomical entity' is not consistent with the intent, we can readily fix this. An 'anatomical entity' should be part of a 'cellular organism'. Most anatomists would likely argue that viral parts are not anatomical entities, though this poses less of an issue for misuse than the organism one above. Will make a ticket.

There are many 'organism' classes in the obo library, we should indeed clean them up, they all mean slightly different things in different ontologies: http://www.ontobee.org/search?ontology=&keywords=organism&submit=Search+terms

What is the current QA/clean up process for potential duplicate entities across OBO? We will be happy to help reconcile but would like to do so using standard process or otherwise help make a standard process.

mellybelly referenced this issue in obophenotype/caro Oct 20, 2018
@mellybelly
updating text and logical def for 'anatomical entity'
3daf188
@bpeters42
Copy link
Contributor

This thread is part of an attempt of a cleanup process for duplicate shared entities across OBO. Yes, there are many other ontologies that define 'organism', the one in OBI being by far the most used (26 vs. 16 for CARO in the page you linked to). My sligthly angry tone comes from being asked to drop yet another OBI term that we spent a lot of time getting right. But it sounds like you are willing to fix CARO, and then we will adopt that in the OBO-Core, and eventually use it to replace the OBI term, and others.

To be concrete, everyone please comment if you agree or object to the below. I took OBIs definition, cut out even more of the potentially controversial bits (viruses don't really 'grow'), and added what I think we are really after, namely the presence of DNA or RNA (to exclude prions and other crap)

Label: 'generalized organism'
Elucidation: "A material entity such as an animal, plant, bacteria, virus or viroid that is capable of replicating or reproducing in the right environment, and that includes genetic information encoded in nucleic acids to do so."

Feel free to disagree, but then propose alternative definitions.

@cmungall
Copy link
Contributor

Just to add general noise and confusion, my position is at odds with everyone... but I have extreme confidence we are converging.

I like names used by scientists. I may be in the minority but I favor using "organism" in the broad sense to include viruses.

I am against both boolean names and adjectives like "generalized" or the elsewhere proposed extended organism. I slightly prefer the boolean name (though I really dislike boolean names or disjunctive OWL definitions) as a non-ontologist can intuit the meaning, whereas "generalized" is too much like ontologese (and doesn't really work on the instance level).

Using organism inclusively is simple, intuitive, and reflects a general conceptual unity amongst its subclasses. I don't see a major risk in curator misuse, there is some risk in that scientists who use the term more restrictedly might attack us for this... but show me an area where this isn't the case. I don't think there is an upper level term in any ontology I work on that isn't controversial. We can of course include obo foundry unique labels, usage notes, logical axioms, and synonyms as well of course.

Both OBI and CARO terms are the product of excellent yet poorly rewarded work by multiple expert people volunteering time. Nevertheless we have to test this work against other communities and be open to revisiting things. On the CARO side, I know we started with a very FMA/BFO philosophy mandate and have gradually moved more towards a more humane ontology, and we should be open to further evolution in that direction.

I think there will also be some shuffling of classes around as part of COREification, and we should have modularity in mind rather than ownership.

Re: definition. I am happy to go with either CARO's clade-based definitions or with OBI's direct definition. However, for the clade-based definitions, there need to be some PURLs for the clade individuals themselves. And then we still have to define those PURLs. On balance it seems simpler to define directly.

I like Bjoern's definition. I mildly prefer moving "such as" clauses to definition gloss (and making it stronger: explicitly stating membership). But currently there is nothing in it that firmly excludes either transposons and possibly even somatic cells. I would like to see "genome" in the definition (although this puts us at risk of circularity (and will need exception clauses for chimeras etc)).

@cmungall
Copy link
Contributor

cmungall commented Oct 21, 2018 •
edited
Loading

FWIW:

  is_a GO:0051704 ! multi-organism process
   is_a GO:0044419 ! interspecies interaction between organisms
    is_a GO:0044403 ! symbiosis, encompassing mutualism through parasitism
     is_a GO:0016032 ! viral process ***

If we make the grouping class label something other than "organism" we should give a heads up on the GO tracker and request GO change to either "multi-generalized organism process" or "multi cellular-organism or virus process".

@mellybelly
Copy link

Ok, lots to think about!

Please everyone, we all have put in a lot of volunteer time and it is exactly the goal of OBO to bring together diverse perspectives into a unified view of biology. No one "owns" content, and it is a basic premise of OBO that one is open to evolution of content to support conceptual alignment. I do think the community is not well tracking all of this hard work though, if one is not attending the RO meeting its a bit opaque - just as an aside it would be good to write an update soon to obo-discuss. You might get more helpers that way too (for example I would be happy to compare and contrast any of the organism/anatomy related duplicates to try to come up with consensus and best placement).

Regarding the above.

I too like the names used by scientists. That is why I don't like the use of 'organism' to mean things that most biologists don't consider organisms ;-). TBH the fact that biologists have not really converged on a label for the union class is very likely why NCBI taxon never had it either. We CAN change this now, but how to do so without confusing or offending? Sometimes labels do need to evolve. What we think of now as a gene is not the same as when that term was originally coined. I don't love 'generalized organism' either but at least is is a flag that it means something different the generally assumed definition of organism. I guess I am willing to be overridden, but lets please better understand consequences if we do want to do that as we are just one small community and we will not change a large community of biologists' mind about such things with one OBO label. Perhaps a poll? Would be good to ask NCBI, EOL and other phylogenetic groups what sorts of requests and suggestions they have gotten.

I like Bjoern's definition too, edits in response to Chris' comments:
"A material entity that is prokaryote (e.g. bacteria, archaea), eukaroyte (e.g. animal, plant, fungus), virus, or viroid and is capable of replicating or reproducing in the right environment, and that includes genetic information encoded in nucleic acids to do so."

Not sure about how to include genome.

I would disagree with the GO hierarchy.... In my mind not all viral processes are 'interspecies interaction between organisms' nor 'symbiosis, encompassing mutualism through parasitism'.

@bpeters42
Copy link
Contributor

bpeters42 commented Oct 21, 2018 via email

@mellybelly
Copy link

Also, (cellular) organism is viewed by anatomists as the root of the anatomical partonomy, so is by definition an anatomical entity (though most anatomists don't use the word entity, thats more of an ontological term).

@bpeters42
Copy link
Contributor

bpeters42 commented Oct 21, 2018 via email

@mellybelly
Copy link

There are cellular anatomists, though they largely fall into studies of single-cell or multi-cellular eukaryotes (and not prokaryotes). GO Cellular Component is considered anatomy and therefore has an equivalency axiom in CARO, for example.

@bpeters42
Copy link
Contributor

bpeters42 commented Oct 21, 2018 via email

@zhengj2007
Copy link

I like Melissa's edited definition too. Regarding labels, I'd prefer 'organism'.

@mellybelly
Copy link

@bpeters42 was just thinking of cellular anatomy - cellular structure and physiology. Things like mitochondria and lysosomes are in GO and are considered cellular anatomy, taught in med school anatomy classes, etc. I don't think most anatomists would consider proteins as anatomy, though obviously there are anatomical structures such as gap junctions that are largely protein complexes and there is a fuzzy line there. We do have the 'gross anatomical part' class to address this issue that was requested by OBI, we can revisit this class too. CARO as an anatomical reference ontology does intend to provide upper level axioms for cellular anatomy of any cell.

@bpeters42
Copy link
Contributor

bpeters42 commented Oct 22, 2018 via email

cmungall added a commit to obophenotype/caro that referenced this issue Feb 8, 2019
@cmungall
relabeled organism class to be simply 'organism'
4856489 
See OBOFoundry/COB#6
@cmungall
Copy link
Contributor

cmungall commented Feb 8, 2019

PR to relabel organism in CARO: obophenotype/caro#17

@cmungall cmungall mentioned this issue Feb 8, 2019
Closed
@cmungall
Copy link
Contributor

Some discussion on organism in OBI, not sure how relevant: obi-ontology/obi#994

@bpeters42
Copy link
Contributor

bpeters42 commented Feb 12, 2019 via email

@cmungall
Copy link
Contributor

cmungall commented Feb 12, 2019 via email

beckyjackson pushed a commit that referenced this issue Jul 11, 2019
Merge pull request #6 from cmungall/cjm-alignments
8d56fb3 
Add mappings to OBOs and external vocabularies (sio, wikidata, biotop, etc)
@beckyjackson
Copy link
Collaborator

We have created an 'organism' term with a unique ID, but we are discussing mapping equivalent external terms in #23.

@ramonawalls ramonawalls mentioned this issue Jul 28, 2019
Open
@mellybelly mellybelly mentioned this issue Aug 3, 2019
Open
cmungall added a commit that referenced this issue Aug 19, 2022
@cmungall
Eliminating ncbitaxon:root and instead bridging to true organism subc…
1948d45 
…lasses

**please read closely and read the related issue before commenting**

This PR removed the statement that COB:organism = NCBITaxon:root

Rationale: root includes non-organisms such as samples

See obophenotype/ncbitaxon#10

This PR replaces that link with two subClassOf axioms

 * COB:organism
    * NCBITaxon:Viruses
    * NCBITaxon:cellular organisms

Note that if this were merged, then the union of NCBITaxon and COB would have a lattice at the top. However, it would be trivial to add NCBITaxon:1 to an anti-slim and filter it out, resulting in a tree with COB:organism as the single MRCA of all organism classes

Merging this PR has some advantages such as obviating the need to rename NCBITaxon:1, since we instead simply inject our own true parent. It leaves the structure of labeling of NCBITaxon intact

**Important** please refrain from discussing the definition of organism here, specifically whether it includes viruses. For COB purposes, this is a closed issue. See: #6
This was referenced Aug 19, 2022
Merged
Open
@StroemPhi StroemPhi mentioned this issue Oct 6, 2022
Open
3 tasks
anitacaron added a commit to obophenotype/caro that referenced this issue Oct 20, 2022
@cmungall @matentzn @anitacaron
relabel organism (#17)
1f7d23d 
* saving with new protege, diff introduces reordering but no actual changes

* relabeled organism class to be simply 'organism'
See OBOFoundry/COB#6

* Update caro.owl

Co-authored-by: Nico Matentzoglu <nicolas.matentzoglu@gmail.com>
Co-authored-by: Anita Caron <anitacaron@users.noreply.github.com>
@cmungall cmungall mentioned this issue Jan 25, 2024
Open
4 tasks
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
9 participants
@cmungall @jamesaoverton @mellybelly @mcourtot @zhengj2007 @cstoeckert @nataled @bpeters42 @beckyjackson