Neglected items

This is a list of various items that have come up at meetings and other discussions, which don't seem to have made it onto any development plan or roadmap.

Measurement value literals in OBD (including ranges)

Phenotypes that include a measurement are not stored without the measurement in OBD, for example "count" or "length" phenotypes. These values need to be imported and provided to the web interface. I'm not sure what you mean here. Can you rephrase as to what is currently happening and what you suggest ought to be happening? --Hilmar 18:35, 19 February 2009 (EST)

Measurement input interface supporting ranges for Phenex

Phenex supports input of a single number for measurements. Many publications provide ranges, which may or may not be bounded on both sides ("5-10", "more than 7").

Fully automated dataloader script

The dataloader should run in an automated way on a server. It should be easy to install and run from any computer. I know Cartik is working on this but I'm not sure it's on a roadmap. --jim 15:53, 19 February 2009 (EST)

Polymorphic matrix states in Phenex

Data matrices sometimes use the NEXUS format for specifying polymorphic states. NeXML supports polymorphic state assignments, but Phenex doesn't provide any facilities for working with polymorphic states.

Search for phenotypes in web interface via anatomy part_of

Currently the web interface and data service provide results using the is_a subsumption relation - searching for "bone" will return annotations to any particular bone. However we have not added a way to search for annotations to parts of some anatomical entity. This should already be supported by OBD - needs to be added to our query services and web interface.

Employing homology statements in data queries (via reasoning?)

We need to determine and explicitly document how homology annotations will impact reasoning related to phenotype annotation searches.

"Systematic character" PATO subset

At the SICB 2009 meeting we discussed choosing a set of terms from PATO that would represent the "character attributes" we will standardly use to group phenotypes in a character matrix (in the web interface, for example). This subset should "bisect" the PATO ontology: any term should either be a descendant of, an ancestor of, or be, one and only one term in the systematic character slim. I believe Wasila and Paula have this on their agenda. --jim 16:15, 19 February 2009 (EST)