tag:blogger.com,1999:blog-5980104128798509938.post4499995143894110260..comments2010-04-06T21:05:29.263-07:00Comments on The Semantic Web in Life Sciences research: Modeling the Linnaean taxonomy in OWL: Where do specimens come in?Cartikhttp://www.blogger.com/profile/06141451252596268391noreply@blogger.comBlogger5125tag:blogger.com,1999:blog-5980104128798509938.post-83969618670667541052010-03-30T08:17:36.155-07:002010-03-30T08:17:36.155-07:00Hi Chris,
In this post, I tried to address the re...Hi Chris,<br /><br />In this post, I tried to address the relationship between specimens and evolutionary taxa that they belong to. This is not addressed in Phenoscape as of now, at least not in a formal manner IMHO. I have talked about the relationships between taxa and phenotypes in my last post and put up some OWL/XML syntax for how the "exhibits" relation can be defined in an OWL framework. <br /><br />In this post, I tried to use a rather bastardized syntax to highlight the points I was trying to make. This syntax was neither this nor that. I apologize for my indolence. I'm currently writing everything down in N3 syntax and I'll put these up in my next post coming soon. As always, thanks for your feedback. <br /><br />- CartikCartikhttps://www.blogger.com/profile/06141451252596268391noreply@blogger.comtag:blogger.com,1999:blog-5980104128798509938.post-13669210967962912182010-03-29T22:15:00.264-07:002010-03-29T22:15:00.264-07:00I recommend using Manchester Syntax as a notation....I recommend using Manchester Syntax as a notation. You state that the following axiom is an RDF triple:<br /><br />'Specimen 1' 'has part' 'some(vertebra 1 and hasQuality some sigmoid)' <br /><br />but it's not in a standard RDF syntax, and the OWL would expand to multiple triples.<br /><br />It's not clear to me if this is treating specimen1 as an individual or class.<br /><br />OWL MS makes the semantics absolutely clear; individual:<br /><br />Individual: Specimen1<br />Types: has part some (vertebra 1 and hasQuality some sigmoid) <br /><br />or:<br /><br />Class: Specimen1<br />SubClassOf: has part some (vertebra 1 and hasQuality some sigmoid) <br /><br />Alternatively you could write everything in a readable RDF syntax like turtle. But IMHO this obscures the OWL.Chris Mungallhttps://www.blogger.com/profile/09523281246244696465noreply@blogger.comtag:blogger.com,1999:blog-5980104128798509938.post-37855854519255236992010-03-29T22:09:27.128-07:002010-03-29T22:09:27.128-07:00Hi Cartik
Can you tell me what a "specimen c...Hi Cartik<br /><br />Can you tell me what a "specimen concept" is?<br /><br />I would have thought the natural way to represent a specimen would be an instance that instantiated a TTO (or similar ontology) class. This is provided you have the ontological commitment that TTO represents individual organisms.<br /><br />OK, perhaps some subtleties. Perhaps the specimen is derived_from an instance of the TTO class (or TTO could be defined sufficiently generously to include derived entities but I don't think this is so good).<br /><br />I'm actually not that sure what constitutes a specimen. Fossils? Pieces of bone? Footprints?<br /><br />You also have:<br /><br />'Specimen 1' 'has taxon name' 'Danio rerio' --(2)<br /><br />I'm not clear on the rationale here. Why do we need a 'has taxon name' relation? What is wrong with relating the specimen to the taxon class, and using the annotation properties in TTO?Chris Mungallhttps://www.blogger.com/profile/09523281246244696465noreply@blogger.comtag:blogger.com,1999:blog-5980104128798509938.post-19668316070948158072010-03-29T13:53:32.250-07:002010-03-29T13:53:32.250-07:00Hi Peter,
Thanks for your thoughts. Given the inf...Hi Peter,<br /><br />Thanks for your thoughts. Given the info on valid names and other names, we can subclass the Name concept to be a valid name or otherwise and define a non-reflexive synoym relation between valid names and other names. Thoughts?<br /><br />- CartikCartikhttps://www.blogger.com/profile/06141451252596268391noreply@blogger.comtag:blogger.com,1999:blog-5980104128798509938.post-79520531245021052632010-03-27T10:32:13.860-07:002010-03-27T10:32:13.860-07:00Taxonomic synonymy is somewhat more complex then y...Taxonomic synonymy is somewhat more complex then you indicate here. A taxon under the Zoological Code has a valid name and may have multiple synonyms - names that have been published, but lack the publication priority of the valid name. Thus synonymy is not symmetric, which the OBO treatment of synonyms captures correctly. I'm not sure what a name being (reflexively) a synonym of itself means in a taxonomic context.Midfordhttps://www.blogger.com/profile/00405629154562504536noreply@blogger.com