Interpretation Model

ClinGen members are generating an incredible amount of richly curated interpretations of the clinical effect of genetic variants. The ClinGen Interpretation Model provides a structure and format for exchanging this information, which retains the contextual and supporting information related to the interpretation of the variant.

ClinGen's interpretation model is a profiles of the Monarch Initiative's Scientific Evidence and provenance Information Ontology (SEPIO), an OWL ontology for the description of evidence and provenance that support scientific claims. This ontology supports a number of use cases, but is mainly driven by the need to integrate evidence, and computationally assess the level of evidence for a scientific claim.
An interpretation is a statement about the pathogenicity of a variant, supported by a structured reasoning process applied to evidence; SEPIO's purpose is to capture exactly such structured data.

Benefits of the ClinGen Interpretation Model

  • Captures a rich set of data around an interpretation:
    • What evidence was used in this interpretation
    • How the evidence supported or conflicted with the interpretation
    • Who performed each individual act in the creation of an interpretation
    • The sources of the data used in the interpretation
    • How the explicitly defined criteria were applied to the evidence
    • To what degree the interpretation builds upon previous interpretations
  • Aligns with related community models including:
  • Other features:
    • Flexibility: This model is naturally able to represent a range of interpretation depth, from unsupported assertions to fully evidence-based interpretations, using whatever level of data has been captured.
    • Contains an extensive set of data objects for representing frequently used information
    • Defines a message structure using industry-standard JSON-LD
    • Extensible: Easily transferable to other types of interpretations other interpretation methods
    • Well-documented, including many examples of using the model to encode variant interpretations using all of the ACMG/AMP guidelines

Development by Example

To ensure that the model supports its intended use case to represent clinical interpretation of variants, development has been guided by real-world examples based on the application of the ACMG guidelines. These examples are available in hierarchical form through the individual Entities pages, or in a "flattened" form in the data/flattened directory of our source code repository.

Status

Version 1 of the interpretation model was released June 27, 2018, and allows the description of variant pathogenicity interpretations based on the ACMG/AMP guidelines. Version 2 will incorporate variant interpretations across a broader set of entities and styles, and is currently in the scoping phase.