SNOMED CT to OWL testbed (Deprecated)

A toolkit for generating test RF2 cases for SNOMED CT to OWL converters and a testing kit for conversions.

As of July, 2019, SNOMED International added the OWL Refset -- a set of OWL Functional Syntax Axioms in tabular form that represent a superset of what was previously contained in the combination of the Concept and Stated Relationship tables. This tool no longer generates valid OWL from post July 2019 SNOMED releases.

It should be noted, however, that the OWL Refset does not:

1. Define annotation properties (e.g. rdfs:label, skos:prefLabel, skos:altLabel and skos:definition). The script referenced below does a reasonable job of providing rdfs:label assertions, but adding the rest of the contents of the Description file and Language refset still needs to be developed. (*Note: it is possible that tools now exist to do all of this -- see the SNOMED Github Site)
2. Provide some of the additional functionality contained within -- RF2Filter in particular

In the short term, some may find the OWLRefsetToOWL script useful.

History

• 0.4.0 - Deprecated as no longer applicable
• 0.3.1 - Add additional core metadata concepts to support 2018 release
• 0.3.2 - Add two further metadata concepts and Modules functio

Requirements

• python 3 -- this has been tested with python 3.6.4
• virtualenv -- useful but not absolutely necessary
• A SNOMED CT RF2 release. (US Citizens can download an image from the NLM Site)

Installation

Option 1: Setup as a python 3 virtual environment

(Assumes virtualenv has been installed)

> virtualenv scttoowl -p python3
> . scttoowl/bin/activate
(scttoowl) > pip install SNOMEDToOWL

Option 2: Install SNOMEDToOWL directly

(Assumes that you are running Python 3)

> python --version
 Python 3.x.y
> pip install SNOMEDToOWL

Modules

• SNOMEDToOWL -- Convert an RF2 Snapshot distribution into OWL
• RF2Filter -- extract selected concepts from an RF2 distribution
• CompareRDF -- Compare two RDF/OWL files

Notes on module imports