Learn how easy it is to sync an existing GitHub or Google Code repo to a SourceForge project! See Demo

Close

#393 Unable to use "decreased" terms for XP term composition

phenotype
closed
6
2011-09-20
2011-04-27
Marcus Chibucos
No

Hello! We are creating a cross product ontology to describe microbial phenotypes, by using PATO terms as the intersection genus and GO as discriminating relations. We have a question about selecting the correct PATO term to use to when creating a term that represents an observed decrease in a biological process. The issue, described below, arises because we are following your recommendations for strict equivalence, i.e. not giving a more general XP definition when a pre-composed term is too specific. (In this case, we feel that the PATO terms might be too specific for our needs.)

For example, we want to make a phenotype term called "decreased cell motility" from a PATO term and the GO term "cell motility". We want a general purpose term that is employable by a typical researcher who might use our ontology and who wants to record the observation that motility is lessened, without necessarily distinguishing among the causes of this observation. That is, we want to be able to capture "decreased cell motility" whether it is due to less frequent movement, slower movement, or movement over a shorter distance.

At present, we do not see how to use PATO to this end, although PATO terms are sufficiently granular to distinguish among each of the three underlying causes:

1. less frequent movement - PATO:0002052 decreased occurrence (an occurrence quality)
2. slower movement - PATO:0000911 decreased rate (a rate, which is a physical quality of a process)
3. movement over a shorter distance - possibly PATO:0000499 decreased duration (a duration quality of a process)

But how to characterize the the observation when it arises due to any of the three causes? I suspect that the problem lies in inheritance--there is not a high-level term that characterizes "decreased" by itself. We have considered requesting "decreased" as a PATO term, but are not sure if this is correct. Would such a term be consistent with other PATO terms such as "PATO:0001508 abolished" and "PATO:0001507 disrupted," the former of which is defined as "...inability to accomplish a biological process."? I would suggest as the definition of "decreased": "A quality of a single process inhering in a bearer by virtue of the bearer's decreased ability to accomplish a biological process." (Yes, it's generic, but it gets around the issue of strict equivalence that we are experiencing.)

The main problem with a term such as "decreased" is where to put it in the ontology. I see that "decreased occurrence" is an "occurrence quality", whereas "decreased rate" descends from "physical quality of a process." Where would our broadly defined term go?

I have read your "PATO:XP_Best_Practice" guide, and I wonder whether we could try a different approach, as well. The section "Complex phenotypes with variable number of conditions" describes a grouping class. Using that as a model, could we do something similar for a grouping class of causes of decrease in an observed process? For example:

The reduction of motility (or various other biological processes), for which we want to merely capture that the phenomenon has been observed, can be caused by any three of the following:

* PATO:0002052 ! decreased occurrence
* PATO:0000911 ! decreased rate
* PATO:0000499 ! decreased duration

First we define a grouping class 'Decreased process cause':

[Term]
id: _:Dp_cause
union_of: PATO:0002052 ! decreased occurrence
union_of: PATO:0000911 ! decreased rate
union_of: PATO:0000499 ! decreased duration

Then we construct our term using the EQ model, and indicate that Dp occurs any time 1 of 3 underlying causes is present:

[Term]
id: microbialphenotype:0000000 ! decreased cell motility
intersection_of: _:Dp_cause {minCardinality="1"}
intersection_of: inheres_in GO:0048870 ! cell motility

Adapting your example was somewhat clunky here, because in the example the union_of terms were all from HP, not PATO itself. Hence, I am not sure if this would be an appropriate solution.

Your insights and suggestions are greatly appreciated. Thank you!

Marcus Chibucos

Discussion

  • Chris Mungall
    Chris Mungall
    2011-05-03

    You shouldn't have to define a union class here - we should hone in on your meaning for generic decreased and ensure there is a PATO term or make a new one.

    Cardinality restrictions on a genus have no meaning, OE should prohibit you from doing this but unfortunately it does not.

     
  • So, does PATO have a protocol for proposing such generic terms? We would need decreased and increased for starters, but I foresee potentially a few more. Can you add those two?
    Thank you!

     
  • I propose two new PATO terms.
    1. "decreased process quality" - def. "A process quality that is relatively low."
    2. "increased process quality" - def. "A process quality that is relatively high."

    The reason I request such generic terms can be found in my initial comment. Primarily, "decreased occurrence," "decreased duration," and "decreased rate" share the most common parent at "process quality." I, personally think that "a process quality that is relatively low/high" is not a terribly meaningful definition. However, I am trying to be consistent with the current definitions of "decreased occurrence," "decreased duration," and "decreased rate."

     
    • labels: 948534 --> new term request
    • assigned_to: nobody --> gkoutos
     
  • Dear Marcus,

    thank you very much for you message. I think what suggest above makes sense. I have created 6 general qualities for increased/decreased process/object these are:

    PATO:0002301 ! decreased quality
    - PATO:0002303 ! decreased object quality
    - PATO:0002302 ! decreased process quality
    PATO:0002300 ! increased quality
    -PATO:0002305 ! increased object quality
    -PATO:0002304 ! increased process quality

    These classes are fully defined so you should get the inference you are after.

    Best wishes,

    George

     
    • status: open --> closed