Chris Mungall wrote:
On May 15, 2008, at 4:39 PM, Kei Cheung wrote:
Hi Vipul,
Take the snomed term "Alzheimer's disease" (AD), for example. One
axis may be used to indicate the progression of the disease. For
example, http://www.memorystudy.org/alzheimers_stages.htm lists 7
stages of Alzheimer's disease: no memory loss symptom, very mild,
mild, moderate, moderately severe, severe, and very severe.
In gene ontology (GO), "memory" (or "learning") is a term under
biological process, one may want to give it a modifier such as
"impaired" if we want to use GO in combination with some modifier
ontology to describe an AD phenotype (e.g., "impaired memory").
In fact both pre- and post- coordinated description of phenotypes
such as "impaired memory" are used. PATO is typically used as the
"modifier" ontology (actually an ontology of qualities and
dispositions). The mammalian phenotype ontology used implicitly pre-
coordinated terms, other groups post-coordinate using PATO + other
OBO ontologies. We have retrospectively gone back to the mammalian
phenotype ontology and attached class expressions defining their
terms to make these approaches comparable. Still very much a work in
progess, see:
http://www.bioontology.org/wiki/index.php/PATO:Pre_vs_Post_Coordinating
Thanks for pointing to PATO describing pre- and post-coordination. It
sounds like one can potentially find the same phenotype in pre- and
post-coordinated ontologies. Is there an issue of redundancy?
Also, I noticed in the document to which you pointed the following: "A
phenotype ontology should be orthogonal to an ontology of molecular
function." For the AD phenotype "impaired memory", researchers may want
to explore to understand the underlying molecular mechanism. For
example, there is a "synaptic plasticity and memory" hypothesis studying
the relationship between synaptic plasticity and memory. In GO, there is
a term called "regulation of synaptic activity", such a biological
process and/or its subclasses may be used to annotate at least partly
the phenotype "impaired memory" ...
Just want to give some examples to illustrate the post-coordinated
approach.
Cheers,
-Kei
Kashyap, Vipul wrote:
Kei,
It's interesting that you bring up this notion.
Most of the vocabularies in healthcare tend to be pre-coordinated,
e.g., ICD9,
CPT4.
However there some compositional terminologies such as RxNorm and
LOINC which
has an underlying model with 6 axes. Snomed tries to be a post-
coordinated terminology, but in
reality it is partially
pre-coordinated and partially post-coordinated.
Cheers,
---Vipul
-----Original Message-----
From: [EMAIL PROTECTED] [mailto:public-semweb-
[EMAIL PROTECTED] On Behalf Of Kei Cheung
Sent: Tuesday, May 13, 2008 12:51 PM
To: Alan Ruttenberg
Cc: Matthias Samwald; public-semweb-lifesci@w3.org
Subject: Re: SenseLab note: some updates
Sounds to me like an ontology exercise. For example, pre-
coordinated approach (knowledgebase) vs. post-coordinated approach
(knowledge base)? May not be a good analogy since I'm not an
ontology expert. :-)
However, I wonder to what extent these two ontology approaches
have been (or have not been) applied to the construction of
ontologies involved in the "knowledge base" or "knowledgebase"
that we are talking about here.
-Kei
Alan Ruttenberg wrote:
Knowledge base is generally written as two words. Try a
google fight
on scholar.google.com.
-Alan
On May 13, 2008, at 8:31 AM, Matthias Samwald wrote:
I have made some small updates to the draft SenseLab conversion
document based on feedback from members of the group. See
http://www.w3.org/2001/sw/hcls/notes/senselab/
Some added text is highlighted in green. Several minor changes
are not highlighted.
Cheers,
Matthias Samwald
DERI Galway, Ireland // Semantic Web Company, Austria
http://www.deri.ie/
http://www.semantic-web.at/
The information transmitted in this electronic communication is
intended only
for the person or entity to whom it is addressed and may contain
confidential
and/or privileged material. Any review, retransmission,
dissemination or other
use of or taking of any action in reliance upon this information by
persons or
entities other than the intended recipient is prohibited. If you
received this
information in error, please contact the Compliance HelpLine at
800-856-1983 and
properly dispose of this information.
