Variant Knowledge Management Software for Labs

When an analyst leaves, their reasoning leaves too

A lab's most valuable asset is its accumulated clinical judgment: the classifications it has made, the evidence behind them, and the reclassification history that goes with them. Inside a closed vendor platform, that knowledge cannot be queried, migrated cleanly, or used as a foundation for AI, and when ClinVar updates, nobody finds out until a clinician asks.

This is not a rare edge case. Variant evidence changes constantly, and variants of uncertain significance are regularly reclassified, so a lab that cannot surface its own precedents ends up repeating interpretation work it has already done.

What GENVAR includes

GENVAR arrives as a working knowledge repository that a lab shapes to its own classification practice.

• A GA4GH VRS canonical store, so the same variant links to one record across every workflow
• Versioned classification events, where a reclassification creates a new version and never overwrites the old
• An evidence graph that draws on ClinVar, gnomAD, and OMIM
• Semantic search across the full interpretation history, powered by a PGVector co-pilot
• ClinVar diff monitoring that surfaces reclassification candidates before a clinician asks
• A repository structured so the lab's own data can ground and train AI assistance

What's ready to run, and what the lab tailors

The canonical store, the versioning model, and the semantic search already exist, so a project starts around the halfway mark rather than at zero.

Because that foundation is in place, NonStop estimates a build reaches production 40 to 50 percent faster than starting from scratch, and the lab owns the result when the engagement ends.

Underneath interpretation, as the lab's memory

GENVAR is the canonical record that every interpretation draws on and writes back to. It gives the interpretation layer its precedents and keeps the institutional history searchable.

From hunting for evidence, down to seconds

* Figures are NonStop estimates and vary by engagement scope.

Because reanalysis of existing data steadily improves diagnostic yield over time, a system that flags variants for re-review turns a forgotten task into recovered diagnoses.

Frequently Asked Questions