Poster Number
9
Poster Title
Cat-VRS After 1.0: Constraint-Driven Evolution of Categorical Variant Representation
Authors
The GKS Categorical Variation Working Group
Abstract
The Categorical Variation Representation Specification (Cat-VRS) defines a unified, constraint-based model for representing categories of genomic variants such as canonical alleles, copy number variants, and protein consequence variants. Released in June 2025, Cat-VRS v1.0.0 offers a strong foundation for genomic knowledge tools and resources that need to describe, search, and exchange categorical variants in a precise, extensible, and interoperable way.
The core innovation of Cat-VRS lies in its constraint model, which provides both generality and modularity. Constraints specify the definitional features of a given categorical variant class (catvar), enabling expressive, harmonizable representation across many variant types, while preserving the flexibility of a "bring-your-own-ontology" style of integration through explicit mappings. In addition to enabling consistent data representation across knowledgebases, constraints also support novel search paradigms: variant class matching can now be performed using logical criteria, rather than imprecise string matching alone.
With the release of version 1.0 in June 2025, adoption momentum has been building. On the knowledge producer side, multiple registered implementers including ClinVar, the Molecular Oncology Almanac, and the VICC MetaKB are incorporating Cat-VRS into their knowledge models. Meanwhile, early knowledge users are exploring pilot use cases, including clinical trial and therapy matching workflows that aim to exploit the precision of constraint-based search.
With a stable core and growing adoption, the Cat-VRS working group is now focused on building out the Cat-VRS constraint system to support more kinds of catvars. We invite new adopters to co-develop the specific constraints they need, whether to capture protein function, fusions and transpositions, systemic markers, or beyond. This ongoing work stands to benefit all users by expanding the scope of Cat-VRS in a principled, interoperable way—driven by real-world needs and shared across the community.
The core innovation of Cat-VRS lies in its constraint model, which provides both generality and modularity. Constraints specify the definitional features of a given categorical variant class (catvar), enabling expressive, harmonizable representation across many variant types, while preserving the flexibility of a "bring-your-own-ontology" style of integration through explicit mappings. In addition to enabling consistent data representation across knowledgebases, constraints also support novel search paradigms: variant class matching can now be performed using logical criteria, rather than imprecise string matching alone.
With the release of version 1.0 in June 2025, adoption momentum has been building. On the knowledge producer side, multiple registered implementers including ClinVar, the Molecular Oncology Almanac, and the VICC MetaKB are incorporating Cat-VRS into their knowledge models. Meanwhile, early knowledge users are exploring pilot use cases, including clinical trial and therapy matching workflows that aim to exploit the precision of constraint-based search.
With a stable core and growing adoption, the Cat-VRS working group is now focused on building out the Cat-VRS constraint system to support more kinds of catvars. We invite new adopters to co-develop the specific constraints they need, whether to capture protein function, fusions and transpositions, systemic markers, or beyond. This ongoing work stands to benefit all users by expanding the scope of Cat-VRS in a principled, interoperable way—driven by real-world needs and shared across the community.
Digital Poster