OBO Release ManagerGUI Documentation
The release manager takes as input an editors version of an ontology and creates a number of release versions in both obo and owl format.
These versions currently include:
- The main version, with additional non-redundant links added via an OWL reasoner (i.e. classified in advance).
- A simple version, which corresponds to the main version with any imported or "mireoted" classes removed
- A version that has not had been reasoned over - the non-classified version
This will help to streamline the process of releasing an OBO ontology (or OWL ontology) by its editors.
The tool also takes care of adding versioning metadata.
In the future it will provide other tasks such as modularization
The application is organized into three tabs (Input/Output, Advanced, Logs) and the 'Make Release'-Button.
[After selecting the input ontology file, pressing this button will start the release process. The view will switch to the Logs-tab. When the release process manager is finished a message box will appear.
This tab represents the minimal number of information required to run the release manager.
-
Input ontology file The ontology to release. This is tested for OBO ontology files, but it also accepts OWL files. However the OWL files are required to be compatible with the OBO2OWL conversion standard. Be sure that the input obo file
- has format-version: 1.2 or higher
- has an ontology: tag with a value that is the same as the ID-space in lowercase (e.g. cl, go)
-
Output folder The folder in which the release manager will create the various ontology files and sub-folders.
This tab contains advanced features and is configured with default values which are reasonable for the main use case.
The core options are described below:
-
Make original (Default: true) Create also the original ontology files in OBO and OWL format. This process does NOT use reasoning and creates the non-classified version of the ontology. The output files generated for ontology {{{X}}} for this option:
- {{{X-non-classified.obo}}}
- {{{X-non-classified.owl}}}
-
Simple (Default: true) The output files generated for ontology {{{X}}} for this option:
- {{{X-simple.obo}}}
- {{{X-simple.owl}}}
-
Allow Overwrite (Default: false) The release manager will only overwrite files without an message or error, if this option is set to true. If this option is set to false, then the user will be asked for permission to overwrite files during the conversion.
-
Reasoner (default: HermiT)
- HermiT
- ELK
- Pellet
- JCel
The Logs-tab contains the logged information of the release manager during the process of making the release. This includes:
- INFO Information about the progress of the application, including a finished message.
- WARN Warnings encountered during the conversion
- ERROR Errors during the release process, which may stop the release from progressing.
On completion you will see something like this:
If the default options are selected, the tool will produce 6 fixes - 3 versions in 2 formats. For example, for the cell ontology, the following 6 files are created:
| cl.obo | cl.owl |
| cl-simple.obo | cl-simple.owl |
| cl-non-classified.obo | cl-non-classified.owl |
Your release directory will look something like this:
[The non-classified file should correspond to the input file. This is prior to reasoning, i.e. it is minimally asserted.]
The main file will include non-redundant additional assertions inferred by the reasoner. Obo format consumers can use this file to perform basic graph traversal operations, as the is_a and relationships should be complete.
The simple file should have all imports and references to external ontologies stripped out. Many obo consumers will want this file. If the source ontology has no imports or MIREOTed terms, the simple file will be identical with the main file.
You can now set up PURLs to point to these, following the guidelines laid out in the OBO Foundry ID Policy document http://obofoundry.org/id-policy.html|ID Policy Document.
The following will be added in future version
- Exporting of subsets
- Support for shortcut relations
- Support for macro-xref expansion (for connecting species ontologies to multi-species ontologies)
- If the input file is obo, make sure obof1.2 or obof1.4 is used
- For obo files, make sure an ontology: directive is present in the header, AND this is identical to the ID-space turned to lowercase (E.g. "go" for the gene ontology)
- If the source is an OWL file, make sure the correct vocabularies are used for annotation properties