Open GoogleCodeExporter opened 9 years ago
Make sure the Infectious Disease Ontology doesn't have this one covered already.
Original comment by hoga...@gmail.com
on 9 Nov 2009 at 6:05
This term will be added in OGMS v0.3
Original comment by albertgo...@gmail.com
on 24 Nov 2009 at 10:16
OBI defines population (OBI_0000181) as "a population is a collection of
individuals
from the same taxonomic class living, counted or sampled at a particular site
or in a
particular area"
Original comment by mcour...@gmail.com
on 1 Dec 2009 at 8:27
We may be able to MIREOT to the OBI definition, but I think there are a few
issues:
(1) organism populations exist even if no one counts or samples them (although I
understand why you want to say this for OBI)...there is a population of people
with
swine flu distinct from the sampled population.
(2) I am a member of the population of the USA, but I can be living overseas and
still be considered a member of that population
(3) there may not be well defined taxonomic classes for certain bacteria (where
at
best we can only cluster them into clonal complexes)
(4) [minor issue] 'individuals' may be too broad
(5) [minor issue] how does 'particular site' differ from 'particular area' (in
terms
of spatial regions)
Just FYI, we need 'organism population' in IDO for very different reasons...to
define
infections as organism populations of infectious organisms, and to define host
and
pathogen populations in epidemiological settings.
Original comment by albertgo...@gmail.com
on 1 Dec 2009 at 8:45
I agree with Albert, that populations exist even if not counted.
This issue and issue #17 raise the question of where certain terms belong, so
to speak.
I appreciate the pioneering work done on OBI, and that folks working on OBI had
to
create high-level terms because of the absence of something like OGMS. And to
the
extent that folks have been using OBI ids for organism population and healthcare
provider, we also need to be mindful of that.
Nevertheless, I would suggest that these terms belong in OGMS, and that the
question
then is whether it is worth the effort to move them there.
I certainly don't want to fight turf wars and won't press the issue any further
than
this, but I just to make sure that terms are in the appropriate place. I don't
know
what the OBO Foundry official position is on issues like this, either.
Original comment by hoga...@gmail.com
on 3 Dec 2009 at 1:51
I agree with #1, populations exist even if not counted. I believe this is an
issue
with the OBI definition, I am not sure OBI developers intended to restrict
this. As
you know we had issues with definition of population/aggregate which lead to the
creation of the MaterialEntity class, so it may very well be that those are not
very
up-to-date. I will email both lists and we'll see.
I am not sure about #2 (population of USA living abroad) that sounds to me like
an
other meaning of the word population, linked to social/legal status, so maybe
more
like a role - it resembles "citizen" for example.
I agree with your other points.
Regarding the need in IDO to define populations of infectious organisms, again,
I am
not sure it is contradictory with the intent of OBI.
Regarding where do terms live, I am happy either way, with the aim of not
creating
duplicate terms.
Original comment by mcour...@gmail.com
on 3 Dec 2009 at 3:45
Original comment by albertgo...@gmail.com
on 4 Dec 2009 at 5:42
Don't close this issue until we have worked out a common definition with OBI
Original comment by alanruttenberg@gmail.com
on 7 Dec 2009 at 4:57
the relevant IDO terms/definitions are:
organism population: an aggregate of organisms
- comment: The aggregate of organisms may be delineated by spatio-temporal
proximity or by demographic
criteria such as age.
simple organism population: An organism population that has as constituents
organisms of the same type.
mixed organism population: An organism population that has as constituents
organisms of different types.
i agree with albert's points above. in addition, i think these terms, as well
as other general terms such as
organism, belong in an ontology that is more general then OGMS, an ontology
that covers general biology
terms that are not specific to medicine. it seems OGMS might have subtypes
that are specific to medicine,
certain types of patient populations perhaps, and OBI might have terms that are
specific to investigations, like
sampled or counted populations/groups. in IDO, we have specific terms like
infectious organism population,
normal resident microbiota, etc.
Original comment by lindsay....@utsouthwestern.edu
on 2 Jan 2010 at 5:48
Definitions that make use of the construct "of the same type" don't work,
because all entities are .... entities. So
everything is of the the same type, at least for one type. "different types" is
similarly problematic.
Original comment by alanruttenberg@gmail.com
on 7 Jan 2010 at 7:10
I think in this case we could just say "...of the same type of organism". I
always
parse the natural language phrase "...of the same type" to mean of the same
immediate
type in the ontology, without going up (for example) as far as entity, but I
guess we
could always be more precise.
Original comment by albertgo...@gmail.com
on 7 Jan 2010 at 5:55
same type of organism has the same problem. All subtypes of organism are of
type ... organism. I.e. the same
type.
"Immediate type" isn't something that can be counted on because it isn't stable
in the face of ontology
evolution that makes finer distinctions as science progresses. So I do think we
should figure out some better
way of expressing this.
I volunteer Barry. :)
Original comment by alanruttenberg@gmail.com
on 7 Jan 2010 at 7:18
Alan...I don't buy that as an argument for not using "immediate type". I think
it is
well understood that as science progresses that ontologies need to change (and
this
may change what we understand to be immediate types)...but the fact that 'fungal
organism population' hinges on 'fungus' makes it a fairly stable type (at least
more
stable than types that talk about universals like higgs bosons).
Philosophers like to talk about the case of jade, which was thought to be a
universal, but after some scientific discovery, what we were calling 'jade' was
actually two minerals 'jadite' and 'nephrite' (see
http://en.wikipedia.org/wiki/Jade )
At that point all ontologies and logical statements about jade had to
change...including implicit references like "immediate type". IMO, we should
never
expect our ontologies aren't going to change in the face of scientific
discovery.
Original comment by albertgo...@gmail.com
on 8 Jan 2010 at 5:14
I agree with Alan. There is no problem in defining 'fungal population' as a
population whose members are fungi. There is a problem in defining 'simple
organism
population' as being homogenous in some way to exclude e.g. a group of fungi and
zebras being considered one population.
Original comment by bjoern.p...@gmail.com
on 8 Jan 2010 at 5:53
When you write "a fungal organism population is an organism population whose
members are organisms of type fungus" and you add a
subtype of the type fungus, the extension of the class does not change.
When you write "a fungal subtype organism population is an organism population
whose members are organisms of the same immediate
type of fungus", the extension does change if you add a subclass.
Suppose you have
Fungus
Fungus type A
Fungus type B
Suppose you have an instance of fungal subtype organism population (call this
instance fsp) all of whose members are fugus type A.
Now suppose you further subdivide into a partition
Fungus type A
Fungus type A1
Fungus type A2
and find that the members of fsp are type A1 and some type A2.
Now the instance fsp is no longer and instance of fungal subtype organism
population.
This kind of change, due solely to adding of detail, isn't acceptable. It
violates monotonicity, for one, and *not for any good reason*.
The definition of fungal organism population doesn't suffer this problem.
In the case of jade, not all statements needed to be changed. That jade is
green and solid remains true. Each statement of the form: "This
little buddha is made of jade" remains true. etc.
-Alan
Original comment by alanruttenberg@gmail.com
on 8 Jan 2010 at 5:54
That is *not* how I was proposing to use immediate type...not subtypes like the
second reading. It was actually more like the first reading
"a fungal organism population is an organism population whose members are
organisms
of type fungus"
where you overwrite 'immediate type' with 'fungus'...this is pretty much what
you are
doing with obi:population...you just use taxonomic class instead of immediate
type(OBI_0000181). If 'taxonomic class' is the silver bullet that solves this
problem then I am happy to accept it.
Is there really a confusion that a fungus population is a population of fungus
:-)?
Original comment by albertgo...@gmail.com
on 8 Jan 2010 at 6:08
Also bjoern, how is it that I understand your use of the word "homogenous" in
your
comment if I don't understand "of the same type"? Do you have a definition for
homogenous that doesn't mention same type?
Original comment by albertgo...@gmail.com
on 8 Jan 2010 at 6:11
FYI...sortal logic tries to deal with expressions containing "...the same X
as...",
if we don't have "the same type as" as a (meta)ontology relationship available,
then
we need to do a lot of additional work that likely goes beyond the scope of
OGMS...
http://plato.stanford.edu/entries/sortals/
Original comment by albertgo...@gmail.com
on 8 Jan 2010 at 6:30
My point is that even if we did have "same type as" as a (meta)ontology
relationship available it would be a bad
idea to use it.
Original comment by alanruttenberg@gmail.com
on 8 Jan 2010 at 6:51
I think there is a misunderstanding here, and I hope it is not me. I thought the
problem raised by Alan is that it is not possible to distinguish
simple organism population and mixed organism population as defined in comment
9, as
something may become 'mixed' as we learn more.
None of us has a problem with 'fungal organism population'.
Alan (and I) would have a problem with 'simple fungal organism population'
defined as
having members of the same type of fungus.
Original comment by bjoern.p...@gmail.com
on 8 Jan 2010 at 6:51
I understand it as Bjoern does.
Original comment by alanruttenberg@gmail.com
on 8 Jan 2010 at 6:57
Thanks for the elucidation Bjoern,
We don't intend 'simple fungal organism population', but rather 'fungal organism
population', 'bacterial organism population', etc...to define 'infection', a
central
term in the infectious disease ontology.
Original comment by albertgo...@gmail.com
on 8 Jan 2010 at 8:09
If the 'simple organism population' will not be used, we are fine.
That leaves the definition of 'organism population'=def: an aggregate of
organisms.
As long as everyone is fine with this including the set of me, a zebra and a
fungus,
it will work. It does for me. Anything stronger would have to be subtyped.
Original comment by bjoern.p...@gmail.com
on 8 Jan 2010 at 8:25
I see the problem with "of the same type", but I don't see a way around it. I
am open to alternative suggestions.
The problem comes when distinguishing "regular" and
mixed/complex/polyorganismal infections/diseases. I
can't think of a way to do this without referring to the fact that there are
infectious organisms all of the same
type in one case and of different types in the other. There is a large
community of researchers who study the
ways in which these types of infections are different from simple
co-infections, so we are making a distinction in
IDO between these types of infections and multiple distinct infections that
happen to be in the same host
organism.
Original comment by lindsay....@utsouthwestern.edu
on 8 Jan 2010 at 8:51
maybe you guys can suggest another way to make this distinction.
Original comment by lindsay....@utsouthwestern.edu
on 8 Jan 2010 at 8:54
re the use of "taxonomic class" instead ... to me it seems that "same taxonomic
class" has exactly the same
problem as "same type" (if i correctly understand the complaint). any two
types have some supertype in
common, and any two taxonomic classes have some parent class in common.
in many cases, we could potentially solve the problem by specifying at which
taxonomic rank the organisms have
to be of the same taxonomic class (ie species).
this doesn't help us with the cases where we are not talking about organisms
though.
Original comment by lindsay....@utsouthwestern.edu
on 7 Feb 2010 at 5:05
I think that would work Lindsey. 'mixed species population' vs. 'single species
population'. This could work as the taxonomic rank 'species' defines at what
'granularity cutoff' two organisms are considered the same even if we later
learn
about additional differences. Similarly, 'mixed genus population', 'mixed strain
population' etc. could be defined.
This does require though that we manage to tie taxonomic ranks to the NCBI
taxonomy
which I believe isn't fully worked out.
When would we not talk about organisms for populations?
- Bjoern
Original comment by bjoern.p...@gmail.com
on 8 Feb 2010 at 4:05
good idea. i will try that.
i guess we would always be talking about organisms when talking about
populations, but there are cases where
we use "of the same type" or "of a different type" where we are not talking
about populations or organisms. for
example, in IDO, we refer to replication where we want to say that the output
is of the same type as one of the
participants. we want this to apply to things other than organisms. (because
you wanted us to include prions!!) i
think there are other non-organism examples as well. but using taxonomic rank,
if we can figure out a way to
tie this to NCBI taxonomy as you mention, might well solve all the organism
cases.
Original comment by lindsay....@utsouthwestern.edu
on 10 Feb 2010 at 12:13
Bjoern makes an important point about the term 'population' being used only in
the context of organisms. Does
this make the use of 'organism' in 'organism population' unnecessary? Perhaps
this clarification should be
mentioned in the comments area.
Do we really have to worry about 'prions' when trying to define 'population'?
As suggested earlier, if '(organism)
population=def: an aggregate of organisms' is taken as an operational
definition, then prions would be
automatically included if they are classified as organisms. The issue would be
differed to where it truly belongs.
Original comment by sivaram....@gmail.com
on 22 Feb 2010 at 10:11
Reviving this old thread.
Lindsay Cowell has pointed out that IDO uses the definition
organism population = An aggregate of organisms of the same Species.
and OGMS uses the definition:
organism population = An aggregate of organisms of the same type.
OBI uses the definition:
population = a population is a collection of individuals from the same
taxonomic class living, counted or sampled at a particular site or in a
particular area
(Note: the issue I raised earlier in the thread about a population not needing
to be counted or sampled to exist is no longer a problem because of the
disjuntion 'living, counted OR sampled')
All three definitions suffer from the 'of the same X' problem that Alan
originally pointed out, but perhaps the IDO or OBI definitions are less bad
because they restricts the ontological type X is. As such, I think we need to
change 'of the same type' in OGMS to 'of the same Species' (per IDO), or 'of
the same taxonomic class' (per OBI).
Which of these is preferable?
Original comment by albertgo...@gmail.com
on 18 Sep 2012 at 7:43
Original issue reported on code.google.com by
albertgo...@gmail.com
on 30 Sep 2009 at 2:18