disorder

[GoogleCodeExporter]

Peter Robinson said: 
2) Disorder
We spoke about this in San Francisco and agreed to disagree for the moment.
I think that "A causally relatively isolated combination of physical
components" is good, but the question of "maximal" creates a lot of
difficulties. In the tumor example with multistage progression (see for
instance Figure 1 of this article
<http://www.pnas.org/content/99/23/15095.long>http://www.pnas.org/content/99/23/
15095.long
>), it is very hard to say what the actual disorder is if we regard the
tumor as a whole as the disorder. I think that the disorder is in fact a 
DNA mutation in a tumor-relevant gene, leading to a disease which is in
fact a predisposition for the cell (and the clone of cells that come from
it) to grow somewhat faster than normal.  When a mutation in a second
tumor-relevant gene occurs in one of the daughter cells of the first cell,
then this is a second disorder that causes a second disease (the daughter
cells of this clone will grow somewhat more). This can continue for some
time until there a clones of cells whose disorder is conceived to be the
union of all tumor-relevant mutations, or which have multiple disorders
(many mutations) and whose disease is metastatic cancer. I think this kind
of "onion" model with multiple layers of disorder/disease is more natural
and requires much less arbitrariness to draw the line between
disorder/disposition/disease.

Original issue reported on code.google.com by albertgo...@gmail.com on 27 Jul 2009 at 3:36

[GoogleCodeExporter]

Lindsay Cowell, Barry Smith, and Albert Goldfain (at a meeting in Duke 9/24/09) 
have
suggested a new definition of disorder:

Disorder =def A disorder is a material entity which is clinically abnormal and 
which
is a fiat object part of the whole formed by: (1) an organism, (2) material 
entities
adhering* to the organism, and (3) any material entities located within the 
convex
hull of the organism (e.g., the bloodstream, the gut, the lungs).

Comment: The relation adhering* is taken to be the transitive closure of the 
adhering
relation, i.e., if x1 adheres to x2 adheres to x3 ….  adheres to xk adheres 
to the
organism, then all of x1, x2, …, xk adhere* to the organism.   Things like 
“shortage
of oxygen” are not disorders, rather, it is a disordered lung (with a reduced
capacity to hold oxygen) that is the disorder and which leads to pathological
processes.   Things like fluid in the lungs are disorders (via clause 3).  

Comment2: Perhaps RO's "adjacent_to" is better than "adheres"

Original comment by albertgo...@gmail.com on 30 Sep 2009 at 2:15

[GoogleCodeExporter]

Is a "genetic disorder" an ogms:disorder?

Original comment by alanruttenberg@gmail.com on 1 Oct 2009 at 12:17

[GoogleCodeExporter]

@alanruttenberg: yes, a genetic disorder is an ogms:disorder.  Do you have a
counterexample to the proposed definition?

Original comment by albertgo...@gmail.com on 1 Oct 2009 at 12:23

[GoogleCodeExporter]

I am concerned with being able to delineate what sorts of genetic mutations are 
disorders. An ovum is formed 
with some variation of the genome from others. How do we know which such 
changes are genetic disorders? Or 
does the genetic disorder only retroactively understood to be such when there 
is a "clinically abnormal" 
manifestation?

Original comment by alanruttenberg@gmail.com on 1 Oct 2009 at 12:28

[GoogleCodeExporter]

in response to alan, comment 4: we are pushing a lot into "clinically 
abnormal".  when you ask whether a 
mutation or particular allele can only be identified as a genetic disorder 
retrospectively, after there is a 
manifestation of disease, are you asking at the level of instances or at the 
level of types?  discovery is 
retrospective in that it usually requires something like the detection of an 
association between a particular 
genotype and a particular phenotype by studying individuals in whom there is a 
clinically abnormal 
manifestation.  after the association is known, the corresponding genetic 
disorder (type) can be defined 
structurally, and instances can be identified through genotyping.  is that 
kind-of what you are getting at?

Original comment by lindsay....@utsouthwestern.edu on 5 Oct 2009 at 4:53

[GoogleCodeExporter]

I am asking at the level of instances, since by my reckoning, all type level 
relations are based on quantification 
(or possibly other operators) over instance level relations. 

I concur that *some* genetic disorders (all of which are structural) are 
sure-fire predictors of phenotype. 

However many variations very connected to environment and there may be no 
clinical manifestation. Consider a 
variation that has the consequence of making the person more likely to have a 
worse bout of some particular 
strain of flu (perhaps not even yet extant). Is this a genetic disorder?

Original comment by alanruttenberg@gmail.com on 6 Oct 2009 at 2:49

[GoogleCodeExporter]

Per my understanding, what you describe as a genetic variation that makes one 
more
likely to contract a particular strain of flu, would be considered a genetic
predisposition to a disease of type X (a pattern, not an actual term), in OGMS 
lingo.

Original comment by hoga...@gmail.com on 6 Oct 2009 at 10:41

[GoogleCodeExporter]

Bill, yes, that is the consequence of the variation - the physical basis. The 
question is, is the variation itself a 
disorder?

Original comment by alanruttenberg@gmail.com on 6 Oct 2009 at 11:41

[GoogleCodeExporter]

Currently, given the definitions in OGMS, the answer is no.  We have not 
defined a
genetic predisposition to a disease of type X as having a disorder as its 
physical
basis, but rather a constitutional genetic abnormality with no requirement to 
exceed
the risk threshold necessary to be clinically abnormal.

I think the question ultimately boils down to: is a causally relatively isolated
combination of physical components that increases one's risk for acquiring a 
disorder
itself a disorder?  And the answer is not necessarily in general, and "no" in 
your
example of genetic predisposition to a particular strain of flu.

Original comment by hoga...@gmail.com on 6 Oct 2009 at 12:25

[GoogleCodeExporter]

Bill. Good. Now what about trisomy? Another genetic variation. Same status?

Original comment by alanruttenberg@gmail.com on 6 Oct 2009 at 12:46

[GoogleCodeExporter]

I think trisomy is a genetic disorder, because the extremely high risk of 
illness,
pain, etc. clearly meets the criteria for clinically abnormal.

But I'm interested to hear why my previous thoughts might lead one to conclude 
otherwise.

Original comment by hoga...@gmail.com on 6 Oct 2009 at 1:05

[GoogleCodeExporter]

perhaps its one of those fuzzy boundary cases. I'm searching for what the 
discriminant for genetic disorder is. 
That certain genetic variation is consider disorder and other not begs for the 
question of how do we discern 
which is which. 

Here the discrimination, following your intuitions, is that the strength of the 
predisposition is much stronger. 

Following the definition:

A disorder is a material entity which is clinically abnormal and which
is a fiat object part of the whole formed by: (1) an organism, (2) material 
entities
adhering* to the organism, and (3) any material entities located within the 
convex
hull of the organism (e.g., the bloodstream, the gut, the lungs).

One might ask, in the newly fertilized ovum, which part is the disorder.

Original comment by alanruttenberg@gmail.com on 6 Oct 2009 at 1:23

[GoogleCodeExporter]

I suppose we should start collecting bits of the definition of "clinically 
abnormal", in part based on your last 
comment.

Original comment by alanruttenberg@gmail.com on 6 Oct 2009 at 1:28

[GoogleCodeExporter]

What about sickle cell trait?

Original comment by sivaram....@gmail.com on 6 Oct 2009 at 1:41

[GoogleCodeExporter]

There have been calls to reclassify sickle cell trait as a disease state, which 
would
make the genetic basis of it a disorder.

Ajayi AA (October 2005). "Should the sickle cell trait be reclassified as a 
disease
state?". Eur. J. Intern. Med. 16 (6): 463. doi:10.1016/j.ejim.2005.02.010. ISSN
0953-6205. PMID 16198915.

But it's a nice boundary, fuzzy, grey case.  And the experts themselves are 
debating it.

Original comment by hoga...@gmail.com on 6 Oct 2009 at 3:09

[GoogleCodeExporter]

Alan, in response to your last note, clinically abnormal is an undefined term in
OGMS.  Per the AMIA translational summit paper, though, here is an explication 
of it:

When we say that some bodily feature of an organism is clinically abnormal, this
signifies that it: (1) is not part of the life plan for an organism of the 
relevant
type (unlike pregnancy or aging), (2) is causally related to an elevated risk 
either
of pain or other feelings of illness, or of death or dysfunction , and (3) is 
such
that the risk exceeds a certain threshold level.

I am not sure about the newly fertilized ovum.  Is it even an organism? (I'm 
thinking
about the 16 days paper).  Does that mean no disorder exists (since it's not an
organism) until 16 days (when we might consider it its own organism)?  

Original comment by hoga...@gmail.com on 6 Oct 2009 at 4:44

[GoogleCodeExporter]

A fertilized ovum, embryo or a fetus is part of the normal physiology and I 
don't
think that we can call it a disorder. Yes, there is a structural change in the 
host
organism (mother) and there is the creation of a new organism (offspring). And, 
this
must be accounted for within the definition of normality. 

What about a prosthetic heart valve? This is not part of the normal physiology. 
There
is also an elevated risk of complications associated with it.  Can we apply the 
same
criteria to define abnormality? 

Original comment by sivaram....@gmail.com on 6 Oct 2009 at 6:11

[GoogleCodeExporter]

I wasn't calling newly fertilized ova, fetuses, or embryos abnormalities.  I was
trying to figure out whether abnormal genes that are part of some instances of 
those
entities are disorders.  So, is trisomy 21 of an embryo a disorder?  Because an
embryo is not an organism (per the 16 days paper).

A prosthetic heart valve is a disorder, per the current OGMS definition of 
disorder.

Original comment by hoga...@gmail.com on 6 Oct 2009 at 6:18

[GoogleCodeExporter]

This creates an interesting situation where a disorder (e.g. mitral stenosis) 
would
be treated by creating yet another disorder i.e. a prosthetic mitral valve! 

If we are describing 'disorder' at the genetic level, how does it matter 
whether we
talk about it in reference to a single cell, or a group of cells, or a specific 
group
of cells called embryo (that is not yet an organism)? If we take a malignant
condition, the disorder must (??) have started in a single cell. 

Perhaps these are two different questions - (1) When is a structural change a
disorder? and (2) Is a disorder of a structure attached to an organism, a 
disorder of
the organism? 

Original comment by sivaram....@gmail.com on 6 Oct 2009 at 7:10

[GoogleCodeExporter]

According to the "16 day paper", I guess the embryo is only a part of
the "mother" organism.

Is that part of the organism subject to the certainty of clinical
abnormality after trisomy? I guess the answer is yes, although part of
the clinical abnormality definition is fully satisfied in the present
situation:

(1) is not part of the life plan for an organism of the relevant
type (unlike pregnancy or aging),

but another part of the definition requires evaluating the future
effects, when the embryo won't be belonging to the organism "mother"
any more:

(2) is causally related to an elevated risk either
of pain or other feelings of illness, or of death or dysfunction ,
and (3) is such that the risk exceeds a certain threshold level.

In addition, this would produce a situation in which the embryo
trisomy is a disorder of the mother (and not of the embryo that's
bearing it), as soon as the embryo is not an organism yet; but then,
as soon as the embryo becomes an organism, that disorder should be
removed from the mother and attached to the offspring. Could this be
source of problems?

In this light, can the "16 days paper" really be assumed as a
rock-solid foundation..?!
If yes, should there be a sub-rule in the definition of disorder
specifically dealing with the disorder of the offspring before it's a
stand-alone organism?

And what about some physical structure of the mother, which is going
to affect only the embryo's fitness, such as immune cells reactive to
the Rhesus (Rh) antigen, that endanger the embryo's survival, but not
the mother's? Is that a disorder of the mother? Is that a disorder at
all?

Original comment by Daniele....@gmail.com on 6 Oct 2009 at 7:23

[GoogleCodeExporter]

Siv, yes, a heart-valve replacement is an exchange of one disorder (with a very 
high
risk of death) for another disorder (with a very low risk of death, but with a
significant risk of other illness and dysfunction, such as thromboembolism,
endocarditis, etc).

Even biological valves have a substantial estimated lifetime risk of serious 
problems:

Prognosis after aortic valve replacement with a bioprosthesis: predictions 
based on
meta-analysis and microsimulation.
AU  Puvimanasinghe JP; Steyerberg EW; Takkenberg JJ; Eijkemans MJ; van 
Herwerden LA;
Bogers AJ; Habbema JD
SO  Circulation 2001 Mar 20;103(11):1535-41.

    BACKGROUND: Bioprostheses are widely used as an aortic valve substitute, but
knowledge about prognosis is still incomplete. The purpose of this study was to
provide insight into the age-related life expectancy and actual risks of 
reoperation
and valve-related events of patients after aortic valve replacement with a 
porcine
bioprosthesis. METHODS AND RESULTS: We conducted a meta-analysis of 9 selected
reports on stented porcine bioprostheses, including 5837 patients with a total
follow-up of 31 874 patient-years. The annual rates of valve thrombosis,
thromboembolism, hemorrhage, and nonstructural dysfunction were 0.03%, 0.87%, 
0.38%,
and 0.38%, respectively. The annual rate of endocarditis was estimated at 0.68% 
for
>6 months of implantation and was 5 times as high during the first 6 months.
Structural valve deterioration was described with a Weibull model that 
incorporated
lower risks for older patients. These estimates were used to parameterize, 
calibrate,
and validate a mathematical microsimulation model. The model was used to 
predict life
expectancy and actual risks of reoperation and valve-related events after
implantation for patients of different ages. For a 65-year-old male, these 
figures
were 11.3 years, 28%, and 47%, respectively. CONCLUSIONS: The combination of
meta-analysis with microsimulation enabled a detailed insight into the prognosis
after aortic valve replacement with a bioprosthesis for patients of different 
ages.
This information will be useful for patient counseling and clinical decision 
making.
It also could serve as a baseline for the evaluation of newer valve types. 

Original comment by hoga...@gmail.com on 6 Oct 2009 at 7:50

[GoogleCodeExporter]

Perhaps I've created a dilemma where none exists, with respect to fertilized 
ova,
embryos, etc.

We can say those things are organisms (separate from the mother) without 
concerning
ourselves as to whether they are human beings.

So, disorders may be part of embryos and the like.

Original comment by hoga...@gmail.com on 6 Oct 2009 at 7:54

[GoogleCodeExporter]

is having a fever having a disorder? What's the part?

Original comment by alanruttenberg@gmail.com on 29 Oct 2009 at 9:10

[GoogleCodeExporter]

well, fever wouldn't be a disorder by the definition. but then what is it?

Original comment by alanruttenberg@gmail.com on 29 Oct 2009 at 9:28

[GoogleCodeExporter]

Barry's Answer:

My view is that fever would be a pathological process
(fever, and running a fever, would be synonymous)

Original comment by Daniele....@gmail.com on 1 Nov 2009 at 11:05

[GoogleCodeExporter]

I think there is some duality, so fever should formalized into two different 
entities:

- fever as a pathological process (a process leading to the increase of body 
temperature)
- fever as a disorder

Concerning the latter, the "physical components clinically abnormal" would be 
all the
body parts that, by having a too high temperature, risk being damaged.

The course of the fever (i.e., a transient fever) would then be useful during 
the
diagnostic process, and I guess that would pertain to fever as a process (not
necessarily pathological).

Original comment by Daniele....@gmail.com on 1 Nov 2009 at 11:12

[GoogleCodeExporter]

What about fever as a 'state' - the result of a temporarily altered 
thermoregulatory activity (process) which in 
itself is a result of an elevated bodily temperature set point.

Original comment by sivaram....@gmail.com on 9 Nov 2009 at 7:49

[GoogleCodeExporter]

Sivaram, in this case, wouldn't 'state' actually equate to OGMS 'bodily 
quality',
i.e. quality of a body component?

If "color and mass" are qualities of a physical component, so should be the
temperature associated to fever.

Original comment by Daniele....@gmail.com on 9 Nov 2009 at 9:46

[GoogleCodeExporter]

That is true and if you look at BioTop merged with BFO, it has biotop:State as 
a descendent of bfo:Quality.

Original comment by sivaram....@gmail.com on 9 Nov 2009 at 9:57

[GoogleCodeExporter]

At the ICBO tutorial on events, processes, etc. we saw a number of upper-level
ontologies that modeled states as occurrents/perdurants.  Although it seems
counterintuitive at first, it makes sense when you analyze it.

The idea is that sitting is state, running is a process.  Both occur over time 
and
are not wholly present at a single point in time (my sitting at my desk since I 
last
sat down until I get up next is still ongoing and not complete).  So, for 
example, in
DOLCE, both State and Process are subtypes of Stative Process, the distinction 
being
that all the parts of the former are of the same subtype as the whole (e.g., all
parts of my sitting are themselves instances of sitting), whereas all parts of 
the
latter are not of the same subtype.

The fever state in its entirety never exists in total at a particular time.  
Thus,
the fever process is your fever state according to this view.  They are the 
same thing.

Original comment by hoga...@gmail.com on 10 Nov 2009 at 12:24

[GoogleCodeExporter]

Indeed the treatment of 'State' as a 'Quality' in BioTop is not intuitive. It 
is far more intuitive to think of 'State' 
as an occurrent as you rightly pointed out. 

If you look at a deep enough level, there is always some process going on. 
However, at a higher level nothing 
may seem to happen. Even when one is sitting still (high level view) there are 
active processes (low level view) 
in play, the out come of which is 'sitting still'. 

Similarly, I think we have Sleep as a state and the related SleepProcess and 
WakeProcess. Infact, both these 
processes are active all the time and it is the balance between them that 
results in either a Sleep or Awake 
states depending on which is dominant.

What about a fever? In most cases when we say fever, we mean raised body 
temperature i.e. a 'state'. A febrile 
response on the other hand would indicate a process of producing a fever.

Original comment by sivaram....@gmail.com on 10 Nov 2009 at 3:31

[GoogleCodeExporter]

I think there is a state of being in a sitting position, and a state of having a
temperature of 102.5 F, that are present in their entirety at a single point in 
time,
and do not have temporal parts.  I would call these "states". This is intuitive 
to me
anyway.

The talk of states as processes is what throws me, not the other way around. 
I've
always thought of states and processes as very different kinds of things.

Obviously there is a febrile process, but at any point in time during a febrile
process there is a temperature quality that is above the normal range, which I 
would
think could be called the febrile state.  I suspect this is why BioTop puts 
states as
kinds of quality.

And the word "fever" is ambiguous and heavily overloaded - I doubt it is 
possible to
ever get clinicians to agree on it being either a continuant or an occurrent. 

Original comment by ksp%ihts...@gtempaccount.com on 10 Nov 2009 at 8:56

[GoogleCodeExporter]

At the instance level, what is an instance of fever state?

If there is one instance of fever state per instant of time, then my fever that
lasted 4 hours is comprised of an infinite number of instances of fever state 
(unless
time is quantized somehow).

I think it's more intuitive to say that there is one instance of fever state 
that
occurred over 4 hours.

However, I do worry that by adding the word "state" to any given quality, one 
could
move every quality over to the occurrent side of taxonomy.  The "blue state" of 
my
car, for example.  My car has a quality of blue.  It does not have a 'blue 
state'.

Or, if blue quality is equivalent to blue state, then what is the distinction 
between
states and other types of qualities?

Original comment by hoga...@gmail.com on 10 Nov 2009 at 1:08

[GoogleCodeExporter]

During the teleconference of 11/19/09, it was decided that the revised 
definition of
disorder should be simplified from

Disorder =def A disorder is a material entity which is clinically abnormal and 
which
is a fiat object part of the whole formed by: (1) an organism, (2) material 
entities
adhering* to the organism, and (3) any material entities located within the 
convex
hull of the organism (e.g., the bloodstream, the gut, the lungs).
(see above http://code.google.com/p/ogms/issues/detail?id=3#c1 )

Clauses (1), (2), and (3) can be removed if we have a term for 'extended 
organism'

Barry Smith has proposed a preliminary definition:
--------------------------------------------------
What follows is based in part on Stephan Schulz,
Philipp Daumke, Barry Smith and Udo Hahn,
“<http://ontology.buffalo.edu/bio/Part&Location.pdf>How
to Distinguish Parthood from Location in
Bioontologies”, Proceedings of AMIA Symposium 2005, Washington DC, 669–673.

We are concerned to define the totality which is
comprised of the organism and all of those
entities inside the interior of the organism,
including microflora, infectious agents attached
to the organism either directly or indirectly,
partially digested food substances, inorganic
molecules within the organism, and so forth.

We define the extended organism as the sum of the
organism and all material entities located within the organism.

This is a first draft definition, which is needed
in order to define what is the bearer of the
disease disposition in the case of an infection.

Comments welcome
BS 

Original comment by albertgo...@gmail.com on 21 Nov 2009 at 6:06

• Added labels: Priority-High
• Removed labels: Priority-Medium

[GoogleCodeExporter]

I like the first draft definition of extended organism.

What about overlapping material entities not entirely located within the 
organism? 
For example, suppose I have an arrow shot into my chest and it comes to rest 
such
that the tail and head of the arrow are not located within me, but just part of 
the
shaft.  Is the entire arrow part of the extended me, or just the parts of the 
arrow
located within me?

How about: an extended organism is the sum of the organism and all material 
entities
located within or overlapping with the organism.

A more relevant medical example might be a nasogastric tube.  A substantial 
length of
the tube is located within the patient, but it extends out of the patient's 
nose.

Original comment by hoga...@gmail.com on 24 Nov 2009 at 9:00

[GoogleCodeExporter]

Incorporating the suggestions of Barry Smith and Bill Hogan we have the current 
proposal:

disorder =def a material entity which is clinically abnormal and part of an 
extended 
organism

extended organism =def an object aggregate consisting of an organism and all 
material entities located within or overlapping the organism.

Issues we have to think about extended organism:
(1) "located within" must cover both disorders which adhere to internal parts 
of the 
organism and disorders suspended in cavities (holes) within the organism

(2) "overlapping" must cover overlapping at a region (e.g., the lodged arrow 
example) and overlapping as contact at an anatomical boundary (e.g., an 
infection on 
the skin...although this is probably a bad example because part of the 
infection has 
likely breached epithelial cells, thus overlapping-a-region)

I think (1) and (2) can be spelled out formally using convex hulls and the 
transitive closures of the part_of and contained_in relations (or compositions 
thereof).  The external part of a lodged arrow would be part of the extended 
organism because it is part of something which overlaps the organism. Whether 
or not 
this external part is a disorder will then be determined by whether it is 
clinically 
abnormal (which is really trickly to say because what is discussed as 
clinically 
normal is usually just the organism and its parts).

In any event, I do see this as progress towards defining disorder broadly 
enough to 
cover all of the strange subtypes.  So, absent any different proposals, I think 
this 
is an incremental change we should make in the next release of OGMS.

Thoughts?

Original comment by albertgo...@gmail.com on 15 Jan 2010 at 6:41

[GoogleCodeExporter]

the overlapping won't cover infections that live outside the surface of the 
skin. Perhaps include something about 
including things that occupy sites formed in part by the organism.

Original comment by alanruttenberg@gmail.com on 15 Jan 2010 at 7:01