Skip to content

VCV Statements

Overview

VCV statement generation aggregates individual SCV (submission-level) classifications into variant-level summary statements. The process combines submissions across conditions and submission levels to produce a hierarchical set of aggregate classification statements for each variant.

The pipeline is implemented across two stored procedures:

  1. gkm_vcv_proc — builds the aggregation tables through a two-layer aggregation hierarchy
  2. gkm_vcv_statement_proc — transforms the aggregation tables into GKS-formatted VCV statements with nested evidence lines, written to gkm_dict_vcv

gkm_dict_vcv is the published product for VCV statements. Null/empty field stripping is applied during bundle assembly at export time (assemble-gkm-dicts.py).

Key Concepts

  • Submission levels — PG, EP, CP, NOCP, NOCL, and FLAG — determine how classifications are combined and whether conflicts are detected. Each submission level aggregates independently; only matching levels can combine. Submission levels are ranked PG > EP > CP > NOCP > NOCL > FLAG, with PG always winning at the top
  • Two-layer hierarchy — the Grouping Layer (Classification Grouping and Priority Grouping steps) aggregates individual SCVs, and the Aggregate Contribution Layer applies winner-takes-all across submission levels to produce final variant-level summaries
  • Single classification format — all VCV statements use classification as the aggregate classification attribute, with an optional conflictingExplanation extension when contributing SCVs disagree. The same single-attribute pattern applies to objectClassification within the proposition

Pipeline Flow

SCV Statements (gkm_dict_scv)
         │
         ▼
┌──────────────────────────────────┐
│  gkm_vcv_proc                    │
│  Classification Grouping         │  Group by variation + group + prop + level [+ tier]
│  Priority Grouping               │  Aggregate tiers within level (somatic only)
│  Aggregate Contribution          │  Winner-takes-all across submission levels
└───────────────┬──────────────────┘
                │
                ▼
┌──────────────────────────────────┐
│  gkm_vcv_statement_proc          │
│  BASE statements (3 steps)       │  Build statement structures from agg tables
│  PRE: inline evidence (3 steps)  │  Propagate evidence through layers
│  FINAL: select all               │  All Aggregate Contribution statements
└───────────────┬──────────────────┘
                │
                ▼
         gkm_dict_vcv

Section Contents

  • Aggregation Rules — submission level logic, classification output formats, review status derivation, and layer hierarchy
  • VCV Procedures — detailed documentation of gkm_vcv_proc and gkm_vcv_statement_proc

Examples

See VCV statement examples in the repository for annotated JSONC examples of germline and somatic aggregate classification statements.