- From: Kashyap, Vipul <VKASHYAP1@PARTNERS.ORG>
- Date: Sat, 9 Sep 2006 17:51:20 -0400
- To: "Alan Ruttenberg" <alanruttenberg@gmail.com>, "Miller, Michael D \(Rosetta\)" <Michael_Miller@Rosettabio.com>
- Cc: "Marco Brandizi" <brandizi@ebi.ac.uk>, <semantic-web@w3.org>, <public-semweb-lifesci@w3.org>
One of the possible modeling options is to express a gene g1 as a subclass
of the Gene class. Consider the model given be Alan below:
> Individual(c1 type(Computation)
> value(geneComputedAsExpressed g1)
> value(geneComputedAsExpressed g2)
> value(geneComputedAsExpressed g3)
> )
How about we express the dual model where:
- c1, c2, c3 are computation instances as above.
- gene g1 is "computed as expressed" in each of these computations.
- each of these could be modeled as instances of gene g1
Individual (g1-c1 type(g1))
Individual (g1-c2 type(g1))
Individual (g1-c3 type(g1))
- Thus g1 subclass of Gene; and g1 = {g1-c1, g1-c2, g1-c3}
Have a limited knowledge of the domain, but seems to me that there will be
multiple "gene expression computations" which could compute the same gene
and that each gene instance is a manifestation of the gene in that computation.
This is a very similar to the situation where, a disease such as Parkinson's
disease should be viewed as a subclass of a Disease Class (as opposed to an
instance of) and instances of a disease are the actual manifestations of a
disease in that person.
Any suggestions which of these approaches is "better" or preferred and why?
Which use cases would justify either of these approaches?
Look forward to the thoughts on this issue.
Cheers,
---Vipul
Received on Saturday, 9 September 2006 21:51:44 UTC