Re: reasoning on gene ontology

Hi Bin,

We use OWL reasoning extensively within the GO consortium, primarily
for automated classification of new terms, and for validation of the
ontology and gene associations. For some more background, see
PMID:20152934 and PMID:20973947 (note that in those papers we used
some custom reasoning tools, but current implementations use OWL
reasoners).

There are not so many applications of non-trivial reasoning over GO
outside the usual ontology development lifecycle use cases (but this
is perhaps true of many ontologies). Most of the analysis and query
tools that are currently in use perform a simple transitive closure
over a subset of the relations, and barely count as
reasoning. However, I expect this situation will slowly change as some
of the more advanced axioms that allow for reasoning are
publicized. There is a lot of work happening now that involves
reasoning over collections of orthogonal OBO library ontologies to
mine gene phenotype associations for example.

I suppose it depends how you define reasoning - I assume here you mean
it in strictly in the sense of deductive reasoning as employed by OWL
and semantic web inference engines, rather than, say, probabilistic
inference.

Regards
Chris

On Apr 5, 2012, at 1:10 PM, bin chen wrote:

> Hi,
>  
> Does anyone perform some reasoning using gene ontology? Any intereting story to share? I recently ran reasoning based on extended GO relations [1], and found that the triples doubled after reasoning. I was also aware of some queries are not able to be performed in relational database. but I havenot found very interesing cases to utlize the inferred results. Any reference? Thanks.
>  
> Best, Bin
> [1]  http://www.geneontology.org/GO.ontology-ext.relations.shtml
> 
> -- 
> Bin Chen
> PHD student, Informatics,
> Indiana University at Bloomington
> http://cheminfo.informatics.indiana.edu/~binchen