Skip to content

Conditions & Traits

Overview

ClinVar submissions (SCVs) reference conditions — diseases, phenotypes, or findings — that a submitter associates with a variant. Internally, ClinVar uses the term "trait" for these condition records, and each submission's traits may differ from the normalized traits that ClinVar assigns at the RCV (aggregate review) level. The gks_scv_condition_proc procedure bridges that gap: it maps each SCV's submitted traits to ClinVar's normalized trait records, builds GKS-compliant condition structures with standardized codings and cross-references, and assembles multi-condition submissions into condition sets.

The procedure executes three logical phases in sequence:

  1. Traits (Step 1) — builds GKS-compliant trait records with standardized codings (MedGen, OMIM, MONDO, HPO, Orphanet, MeSH, EFO)
  2. Condition Mapping (Steps 2–14) — resolves each SCV's submitted traits to ClinVar's normalized RCV traits through a multi-stage matching strategy
  3. Condition Sets (Step 15) — assembles individual conditions into Condition or ConditionSet domain entities for each SCV

These steps do not produce a standalone output file. Instead, the resulting gks_scv_condition_sets table feeds directly into the SCV and RCV record assembly, where conditions become part of the full statement output.

ClinVar Trait Terminology

ClinVar's XML schema uses "trait" where the GKS output uses "condition." The mapping between the two terminologies is:

ClinVar Term GKS Term Description
Trait Condition A single disease, phenotype, or finding
TraitSet ConditionSet A group of traits with a membership operator
TraitMapping — ClinVar's internal record linking an SCV's submitted trait text to a normalized trait
ClinicalAssertionTrait — The submitter's original trait record as submitted to ClinVar

Pipeline Flow

┌──────────────────────────────────────────────────────────────────┐
│  gks_scv_condition_proc                                          │
│                                                                  │
│  Step 1:     Traits (temp_gks_trait)                             │
│  Steps 2–14: Condition Mapping (gks_scv_condition_mapping)       │
│  Step 15:    Condition Sets (gks_scv_condition_sets)             │
│                                                                  │
│  Inputs:  trait, trait_mapping, rcv_mapping,                     │
│           clinical_assertion_trait,                               │
│           clinical_assertion_trait_set                            │
│                                                                  │
│  Outputs: gks_scv_condition_mapping, gks_scv_condition_sets      │
└──────────────────────────┬───────────────────────────────────────┘
                           │
                           ▼
                  SCV Record Assembly

Why Condition Mapping Is Complex

Submitters to ClinVar provide their own trait text and cross-references, which frequently differ from the normalized trait records that ClinVar curators assign at the RCV level. The condition mapping procedure exists to reconcile these differences through a progressive matching strategy:

  • Trait mappings — ClinVar provides explicit mapping records that link an SCV's submitted trait to a normalized trait via name or cross-reference. These are the highest-confidence matches
  • Direct xref matching — when trait mappings are insufficient, the procedure matches SCV traits to RCV traits by comparing submitted cross-references (MedGen, OMIM, MONDO, HPO, Orphanet, MeSH) directly
  • Singleton inference — when only one unmatched SCV trait and one unmatched RCV trait remain for a submission, they are paired by elimination
  • Rogue trait matching — for SCV traits that do not appear in the expected RCV trait set, the procedure searches across all normalized traits by xref IDs, preferred name, and alternate names

See Condition Mapping for the full procedure documentation.

Detailed Documentation