ROBOT script as basis for FMA subset

[cuboideum]

The 'extract' function of the ontology application ROBOT allows to create defined subsets of existing ontologies. Such scripts would help to recreate RDFBones subsets from new versions of the source ontologies.

For the FMA subset we need to create a script that creates a subset that meets the criteria outlined in issue #125.

Information on how to download and install ROBOT can be found here:

https://robot.obolibrary.org/

This is also the startpage of the ROBOT documentation. Specific information about the 'extract' function can be found here:

https://robot.obolibrary.org/extract

The 'extract' function supports several methods. Use the MIREOT method for this task.

[cuboideum]

I wonder if it is possible to extract properties just as well as classes. I had some problems with importing instances while working on the implementation of the Phaleron skeletal inventory template. There, I found out that commands for the extraction of instances need to go before commands for class extraction in the script.

So there seems to be a point in playing around with the code in case that there are problems with property definitions.

[cuboideum]

In addition to the requirements outlined in issue #125, we could also include joints. This implies adding the following class tree:

• 'Anatomical junction', fma5898 (subclass of 'Anatomical cluster', fma49443)
  • 'Joint', fma7490
    • All children below 'Joint'

[15E601]

Do you have further plans for this issue, specifically the script? Unfortunately I encountered some limits to ROBOT that made full automation impossible (lack of disjoint-attributes when importing, in particular) - I do not remember how we addressed this problem, though I believe we simply accepted that the FMA classes are not disjoint.

[cuboideum]

The main problem, as I recall, was that inverse properties where not imported. Is this correct? I could live without the disjoint class statements but hardly without the inverse property statements.