[Ontology-editors] CC organization question for xps (fwd)

[Midori Harris]

just noticed that the list was missed out on the last bit of this thread, 
so forwarding for the record

---------- Forwarded message ----------
Date: Thu, 20 Nov 2008 08:37:48 -0500
From: David Hill <dph at informatics.jax.org>
To: Midori Harris <midori at ebi.ac.uk>
Cc: Tanya Berardini <tberardi at acoma.stanford.edu>,
     Chris Mungall <cjm at berkeleybop.org>
Subject: Re: [Ontology-editors] CC organization question for xps

Yup. I agree. I think this is working out.

David

Midori Harris wrote:
> I'm getting happier and happier. The latest proposal looks very good indeed.
> 
> m
> 
> On Wed, 19 Nov 2008, Tanya Berardini wrote:
> 
>> Yes.  This was the crux of my argument with David the other day when I was
>> trying to push for cellular components (cells) not being anatomical
>> structures.  He convinced me, based on the current contents of FMA, that
>> they are.  I was trying to push for FMA losing cells or AT LEAST cellular
>> component but that's pretty much out of our hands as GO.
>> 
>>> Should we just decide within GO at what level anatomical structures end?
>> 
>> If we want to play nice with FMA (or CARO), maybe not.
>> 
>>> From CARO:
>> 
>> [Term]
>> id: CARO:0000013
>> name: cell
>> def: "Anatomical structure that has as its parts a maximally connected
>> cell compartment surrounded by a plasma membrane." [CARO:MAH]
>> is_a: CARO:0000003 ! anatomical structure
>> xref: GO:0005623
>> 
>> [Term]
>> id: CARO:0000014
>> name: cell component
>> def: "Anatomical structure that is a direct part of the cell." [CARO:MAH]
>> is_a: CARO:0000003 ! anatomical structure
>> relationship: part_of CARO:0000013 ! cell
>> 
>> Running with Midori's last proposal and modifying the part of the
>> graph under dev. process to take into account the
>> anatomical structure devt and morphogenesis terms:
>> 
>> How do we integrate this part of David's comments:
>> 
>>> I also have no real problem with creating the 'anatomical structure X'
>>> terms if we deem that CCs are anatomical structures. It would make the
>>> ontology more complete. Although I would only see it having 2 children,
>>> 'anatomical structure X involved in morphogenesis' and 'cellular
>>> component X'.
>> 
>> I went with 'anatomical structure formation involved in development'
>> vs 'anatomical structure formation involved in
>> morphogenesis'.
>> 
>>
>>  biological process
>>  [i] anatomical structure formation (GO:0048646 or GO:new)
>>  --[i] anatomical structure formation involved in development (GO:new
>> or GO:0048646)
>>  ----[child terms; see below]
>>  --[i] cellular component assembly GO:0022607
>>
>>  developmental process
>>  [i] anatomical structure development
>>  --[p] anatomical structure morphogenesis
>>  ----[p] anatomical structure formation involved in development
>> (GO:new or GO:0048646)
>>  ------[all the current formation-of-macroscopic-anatomical-structure
>> children of GO:0048646 would go here]
>>  [i] anatomical structure formation involved in development (GO:new
>> or GO:0048646)
>>  ----[all the current formation-of-macroscopic-anatomical-structure
>> children of GO:0048646 would go here]
>> 
>> 
>> Yes? Have at it.
>> 
>> 
>> Tanya
>> 
>> 
>> 
>> 
>> On Wed, Nov 19, 2008 at 6:03 AM, Midori Harris <midori at ebi.ac.uk> wrote:
>> 
>>> Exactly! Now I think we're finally on the same page. Eagerly awaiting input
>>> from Chris ...
>>> 
>>> 
>>> m
>>> 
>>> On Wed, 19 Nov 2008, David Hill wrote:
>>>
>>>  I think the first thing we have to decide is whether cellular components
>>>> and cells for that matter are anatomical structures. I think this is
>>>> actually in the realm of OBO. It would make sense to me if anatomy went 
>>>> down
>>>> to the supercellular level, then we would have a cell ontology and a
>>>> subcellular component ontology. I guess the question would be do we want 
>>>> to
>>>> have an anatomy that has anatomy-cell Xproducts and cell-CC Xproducts, or
>>>> even further and anatomy that has anatomy-cell-CC Xproducts. This I think 
>>>> is
>>>> up to OBO. Chris, do you have any thoughts on this? Should we just decide
>>>> within GO at what level anatomical structures end?
>>>> 
>>>> I also have no real problem with creating the 'anatomical structure X'
>>>> terms if we deem that CCs are anatomical structures. It would make the
>>>> ontology more complete. Although I would only see it having 2 children,
>>>> 'anatomical structure X involved in morphogenesis' and 'cellular component
>>>> X'.
>>>> 
>>>> 
>>>> David
>>>> 
>>>> Midori Harris wrote:
>>>> 
>>>>> more comments ...
>>>>> 
>>>>> On Wed, 19 Nov 2008, David Hill wrote:
>>>>>
>>>>>  structure, i.e.
>>>>>>>
>>>>>>>  anatomical structure [some CARO ID, I presume]
>>>>>>>  [i] cellular component GO:0005575
>>>>>>> 
>>>>>>> then it follows logically that
>>>>>>>
>>>>>>>  anatomical structure formation GO:[0048646 or new ID; see below]
>>>>>>>  [i] cellular component assembly GO:0022607
>>>>>>>
>>>>>>>  [snip]
>>>>> 
>>>>>>
>>>>>>  If it's because the existing anatomical structure formation term is_a
>>>>>>> developmental process, we can get around that by either moving or 
>>>>>>> renaming
>>>>>>> GO:0048646 and adding a new term, yielding:
>>>>>>>
>>>>>>>  biological process
>>>>>>>  [i] anatomical structure formation (GO:new)
>>>>>>> 
>>>>>> 
>>>>>> Tanya and I discussed this, but we couldn't think of an anatomical
>>>>>> structure other than a cellular component that was formed in a
>>>>>> non-developmental context. So we thought it would be simpler just to 
>>>>>> leave
>>>>>> the cellular component term as the least specific parent.
>>>>>> 
>>>>> 
>>>>> I guess I don't see what it gains us to leave out CC formation is_a
>>>>> anatomical structure formation (assuming the new, more broadly defined 
>>>>> anat.
>>>>> str. form.), and we'd lose consistency with our claim that CC is_a
>>>>> anatomical structure.
>>>>>
>>>>>  We could also add the term  suggested here.
>>>>>>
>>>>>>   --[i] anatomical structure formation involved in development
>>>>>>> (GO:0048646)
>>>>>>>  ----[child terms; see below]
>>>>>>>  --[i] cellular component assembly GO:0022607
>>>>>>> 
>>>>>> 
>>>>> As you've probably guessed, I'm in favor of using this as long as we say
>>>>> CC is_a anatomical structure ... it's completely consistent with the idea
>>>>> that CCs are the only anatomical structures that are formed in
>>>>> non-developmental contexts.
>>>>>
>>>>>   developmental process
>>>>>>>  [i] anatomical structure formation involved in development (GO:new or
>>>>>>> GO:0048646)
>>>>>>>  --[all the current formation-of-macroscopic-anatomical-structure
>>>>>>> children of GO:0048646 would go here]
>>>>>>> 
>>>>>>> Or we can abandon the notion that a CC is_a anatomical structure (which
>>>>>>> must mean there are some CCs that aren't anatomical structures).
>>>>>>> 
>>>>>> We talked about this. This depends on OBO. In the FMA (Foundational
>>>>>> model of anatomy), everything down to a protein complex is an anatomical
>>>>>> structure. Is the FMA going to be the OBO standard for the 
>>>>>> representation of
>>>>>> anatomy? I
>>>>>> 
>>>>> 
>>>>> I don't know; I've been going on the discussions we (you, me, Chris,
>>>>> Tanya, the other EBI GO editors) have had recently where we've been 
>>>>> saying
>>>>> that we think CC is_a anatomical structure.
>>>>>
>>>>>  have nothing against abandoning the notion, but I talked Tanya out of it
>>>>>> yesterday based only on the FMA.
>>>>>> 
>>>>>>> I'm not sure I feel very strongly whether we keep or drop that idea,
>>>>>>> but I do think that we should either make both statements or neither; 
>>>>>>> it's
>>>>>>> logically inconsistent to have just one.
>>>>>>> 
>>>>>>>
>>>>>>>  my 2c
>>>>>> 
>>>>>> David
>>>>>> 
>>>>>> 
>>>>> m
>>>>> 
>>>> 
>>>> 
>>>> 
>> 
>> 
>>

-- 
David P. Hill, Ph.D.
Bioinformatics Scientist: Ontology Development
Gene Ontology Consortium
The Jackson Laboratory
www.geneontology.org
www.informatics.jax.org