Skip to content

Managing your ODK repository

Updating your ODK repository

Your ODK repositories configuration is managed in src/ontology/gcbo-odk.yaml. The ODK Project Configuration Schema defines all possible parameters that can be used in this config YAML. Once you have made your changes, you can run the following to apply your changes to the repository:

sh run.sh make update_repo

There are a large number of options that can be set to configure your ODK, but we will only discuss a few of them here.

NOTE for Windows users:

You may get a cryptic failure such as Set Illegal Option - if the update script located in src/scripts/update_repo.sh was saved using Windows Line endings. These need to change to unix line endings. In Notepad++, for example, you can click on Edit->EOL Conversion->Unix LF to change this.

Managing imports

You can use the update repository workflow described on this page to perform the following operations to your imports:

  1. Add a new import
  2. Modify an existing import
  3. Remove an import you no longer want
  4. Customise an import

We will discuss all these workflows in the following.

Add new import

To add a new import, you first edit your odk config as described above, adding an id to the product list in the import_group section (for the sake of this example, we assume you already import RO, and your goal is to also import GO):

import_group:
  products:
    - id: ro
    - id: go

Note: our ODK file should only have one import_group which can contain multiple imports (in the products section). Next, you run the update repo workflow to apply these changes. Note that by default, this module is going to be a SLME Bottom module, see here. To change that or customise your module, see section "Customise an import". To finalise the addition of your import, perform the following steps:

  1. Add an import statement to your src/ontology/gcbo-edit.owl file. We suggest to do this using a text editor, by simply copying an existing import declaration and renaming it to the new ontology import, for example as follows:
    ...
    Ontology(<https://github.com/bmir-radx/gcbo.owl>
    Import(<https://github.com/bmir-radx/gcbo/imports/ro_import.owl>)
    Import(<https://github.com/bmir-radx/gcbo/imports/go_import.owl>)
    ...
    
  2. Add your imports redirect to your catalog file src/ontology/catalog-v001.xml, for example:
    <uri name="http://purl.obolibrary.org/obo/gcbo/imports/go_import.owl" uri="imports/go_import.owl"/>
    
  3. Test whether everything is in order:
    1. Refresh your import
    2. Open in your Ontology Editor of choice (Protege) and ensure that the expected terms are imported.

Note: The catalog file src/ontology/catalog-v001.xml has one purpose: redirecting imports from URLs to local files. For example, if you have

Import(<http://purl.obolibrary.org/obo/gcbo/imports/go_import.owl>)

in your editors file (the ontology) and

<uri name="https://github.com/bmir-radx/gcbo/imports/go_import.owl" uri="imports/go_import.owl"/>

in your catalog, tools like robot or Protégé will recognize the statement in the catalog file to redirect the URL http://purl.obolibrary.org/obo/gcbo/imports/go_import.owl to the local file imports/go_import.owl (which is in your src/ontology directory).

Modify an existing import

If you simply wish to refresh your import in light of new terms, see here. If you wish to change the type of your module see section "Customise an import".

Remove an existing import

To remove an existing import, perform the following steps:

  1. remove the import declaration from your src/ontology/gcbo-edit.owl.
  2. remove the id from your src/ontology/gcbo-odk.yaml, eg. - id: go from the list of products in the import_group.
  3. run update repo workflow
  4. delete the associated files manually:
    • src/imports/go_import.owl
    • src/imports/go_terms.txt
  5. Remove the respective entry from the src/ontology/catalog-v001.xml file.

Customise an import

By default, an import module extracted from a source ontology will be a SLME module, see here. There are various options to change the default.

The following change to your repo config (src/ontology/gcbo-odk.yaml) will switch the go import from an SLME module to a simple ROBOT filter module:

import_group:
  products:
    - id: ro
    - id: go
      module_type: filter

A ROBOT filter module is, essentially, importing all external terms declared by your ontology (see here on how to declare external terms to be imported). Note that the filter module does not consider terms/annotations from namespaces other than the base-namespace of the ontology itself. For example, in the example of GO above, only annotations / axioms related to the GO base IRI (http://purl.obolibrary.org/obo/GO_) would be considered. This behaviour can be changed by adding additional base IRIs as follows:

import_group:
  products:
    - id: go
      module_type: filter
      base_iris:
        - http://purl.obolibrary.org/obo/GO_
        - http://purl.obolibrary.org/obo/CL_
        - http://purl.obolibrary.org/obo/BFO

If you wish to customise your import entirely, you can specify your own ROBOT command to do so. To do that, add the following to your repo config (src/ontology/gcbo-odk.yaml):

import_group:
  products:
    - id: ro
    - id: go
      module_type: custom

Now add a new goal in your custom Makefile (src/ontology/gcbo.Makefile, not src/ontology/Makefile).

imports/go_import.owl: mirror/ro.owl imports/ro_terms_combined.txt
    if [ $(IMP) = true ]; then $(ROBOT) query  -i $< --update ../sparql/preprocess-module.ru \
        extract -T imports/ro_terms_combined.txt --force true --individuals exclude --method BOT \
        query --update ../sparql/inject-subset-declaration.ru --update ../sparql/postprocess-module.ru \
        annotate --ontology-iri $(ONTBASE)/$@ $(ANNOTATE_ONTOLOGY_VERSION) --output $@.tmp.owl && mv $@.tmp.owl $@; fi

Now feel free to change this goal to do whatever you wish it to do! It probably makes some sense (albeit not being a strict necessity), to leave most of the goal instead and replace only:

extract -T imports/ro_terms_combined.txt --force true --individuals exclude --method BOT \

to another ROBOT pipeline.

Add a component

A component is an import which belongs to your ontology, e.g. is managed by you and your team.

  1. Open src/ontology/gcbo-odk.yaml
  2. If you dont have it yet, add a new top level section components
  3. Under the components section, add a new section called products. This is where all your components are specified
  4. Under the products section, add a new component, e.g. - filename: mycomp.owl

Example

components:
  products:
    - filename: mycomp.owl

When running sh run.sh make update_repo, a new file src/ontology/components/mycomp.owl will be created which you can edit as you see fit. Typical ways to edit:

  1. Using a ROBOT template to generate the component (see below)
  2. Manually curating the component separately with Protégé or any other editor
  3. Providing a components/mycomp.owl: make target in src/ontology/gcbo.Makefile and provide a custom command to generate the component
    • WARNING: Note that the custom rule to generate the component MUST NOT depend on any other ODK-generated file such as seed files and the like (see issue).
  4. Providing an additional attribute for the component in src/ontology/gcbo-odk.yaml, source, to specify that this component should simply be downloaded from somewhere on the web.

Adding a new component based on a ROBOT template

Since ODK 1.3.2, it is possible to simply link a ROBOT template to a component without having to specify any of the import logic. In order to add a new component that is connected to one or more template files, follow these steps:

  1. Open src/ontology/gcbo-odk.yaml.
  2. Make sure that use_templates: TRUE is set in the global project options. You should also make sure that use_context: TRUE is set in case you are using prefixes in your templates that are not known to robot, such as OMOP:, CPONT: and more. All non-standard prefixes you are using should be added to config/context.json.
  3. Add another component to the products section.
  4. To activate this component to be template-driven, simply say: use_template: TRUE. This will create an empty template for you in the templates directory, which will automatically be processed when recreating the component (e.g. run.bat make recreate-mycomp).
  5. If you want to use more than one component, use the templates field to add as many template names as you wish. ODK will look for them in the src/templates directory.
  6. Advanced: If you want to provide additional processing options, you can use the template_options field. This should be a string with option from robot template. One typical example for additional options you may want to provide is --add-prefixes config/context.json to ensure the prefix map of your context is provided to robot, see above.

Example:

components:
  products:
    - filename: mycomp.owl
      use_template: TRUE
      template_options: --add-prefixes config/context.json
      templates:
        - template1.tsv
        - template2.tsv

Note: if your mirror is particularly large and complex, read this ODK recommendation.