Helix Insight

Classification

Variant classification follows the ACMG/AMP 2015 guidelines implemented through the Bayesian point-based framework (Tavtigian et al. 2018, 2020) with ClinGen SVI calibrated computational predictor thresholds (Pejaver et al. 2022) and SpliceAI integration aligned to ClinGen SVI 2023 recommendations (Walker et al. 2023).

Classification is strictly evidence-based. No machine learning model determines variant pathogenicity. The framework evaluates 28 evidence criteria -- 19 automated, 9 requiring manual curation -- and combines them using calibrated point values to produce one of five standard classifications: Pathogenic, Likely Pathogenic, VUS, Likely Benign, or Benign.

Classification Priority Order

Classification logic is applied in strict priority order. Higher-priority rules are evaluated first, and the first matching rule determines the final classification:

1

BA1 Stand-alone

Allele frequency > 5% is always classified Benign. BA1 cannot be overridden by any other evidence, including ClinVar assertions.

2

High-Confidence Conflict Check

When pathogenic evidence at Strong or Very Strong level conflicts with Strong benign evidence, the variant is flagged for manual review regardless of point total.

3

ClinVar Override

ClinVar classification applied only when no conflicting computational evidence exists. Requires minimum 1 review star. ClinVar VUS does not override computational classification.

4

Bayesian Point System

Each criterion contributes calibrated points. Total determines classification: >= 10 Pathogenic, 6-9 Likely Pathogenic, 0-5 VUS, -1 to -5 Likely Benign, <= -6 Benign.

5

Default

Variants not meeting any rule are classified as Variant of Uncertain Significance (VUS).

Classification Output

Each variant receives one of five ACMG classifications, a list of triggered criteria with evidence strength levels, a Bayesian point total, and a continuous confidence score.

Pathogenic

>= 10 pts

Likely Pathogenic

6-9 pts

VUS

0-5 pts

Likely Benign

-1 to -5 pts

Benign

<= -6 pts

In This Section

For the full methodology with all thresholds and implementation details, see the dedicated Methodology page.