ISF SVN organization

The ISF beta release is composed of a set of files hosted at the “connect-isf” Google Code project [1].  This project uses Subversion (SVN) for its code repository. The following describes the current content of the files and layout in the SVN repository.

Figure 1 shows the SVN directory layout. The top directory is “ontology” and it contains the “arg.owl” OWL file that imports all necessary files (from various subdirectories) to give a full view of the ISF. This directory has the following four subdirectories:

• The “module” folder contains files that cover the main content of the ISF ontology and is described in the next section. 
• The “app-views” directory has application-specific files that import one or more of the application-independent ISF files from the other directories (imports, mireot, module). The “eagle-i” subdirectory contains files specific to the eagle-i application [2] and the “vivo” subdirectory is for the VIVO application [3].
• The “imports” directory contains copies of external ontologies (such as VCARD or Basic Formal Ontology) that are used (through owl:imports directives) by one or more of the other files. This directory is intended to serve as a cache for OWL files that are used in OWL imports so that they can be resolved locally when needed. This also means that this directory could contain an ontology document for a specific ontology but the document returned on the web might be different if it has been recently updated.
• The “mireot” directory is for MIREOTed [4] files.  A MIREOT file is a file that contains some OWL content (classes, properties, annotations, etc.) coming from other existing ontologies (such Ontology for Biomedical Investigations, OBI or the Gene Ontology, GO) that are used in the ISF ontology - but not the whole ontology as in the case of the imports directory. This approach helps avoid having to import full ontologies and the scripted nature of this approach also helps with the maintenance of these files.

Other SVN directories organize ISF related documentation (doc folder), tools, and data.

The tools folder will contain custom scripts that will be developed for the final release of the ontology while the data folder is meant for any related data files such as ISF-compliant RDF data or examples to demonstrate the correct use of the ISF.

Figure 1. The tree structure of the ISF SVN repository

The SVN layout makes a distinction between ontology content that should be reusable and content that is application specific. The final release of the ISF will include some reorganization of these folders with an additional “source” subdirectory that will provide an additional level of organization in addition to the two already described (ontology vs. application). This will be detailed in the final documentation.

The main file that collects all the other files through a owl:import chain is the arg.owl under the ontology folder (this file will be renamed isf.owl in the final release).

All the classes, properties, and individuals in the various files that have an identifier in the form of ARG_XXXXXXX are the entities we have created during the CTSAconnect project to meet some requirements that  couldn’t be met by reusing entities already existing in other ontologies. The prefix ARG stands for Agent, Research resources and Grants, and is a unique prefix for use in the OBO Foundry [5].

Next: ISF modules