search

nmdp-bioinformatics / dash

Data Standards Hackathon for NGS based typing.
GNU Lesser General Public License v3.0
13 stars 13 forks source link

URI contents #11

Open mpresteg opened 10 years ago

mpresteg commented 10 years ago

Part of the beauty of the URI is the flexibility it provides. However, as a consumer of HML (for instance), if I'm interested to follow the consensus-sequence URI, what may I expect to find on the other side? Is it resolvable by a machine? If so, some context (a content type) or something would seem helpful. As a community can we establish this context even if we don't wish to constrain the specification with it?

mpresteg commented 10 years ago

In looking at the consensus-sequence example, there is an implicit 'definition' of what would lay behind the URI (if it were present): targeted-region, sequence and all the elements contained therein.

In the case of the gl-resource URI, this is well defined by the spec: "element from gl-resource.xsd". In the case of glstring URI, this is a string. Implicitly in the syntax of a Genotype List String (as published in Tissue Antigens).

The 'least' defined URI space within the HML spec right now pertains to the raw-reads.

ghost commented 10 years ago

Good point. See also https://github.com/nmdp-bioinformatics/hml/issues/14