Why use it

Provides a universal framework applicable to ALL variant types (CNVs, ROH, regulatory, hypomorphic). Separates functional from clinical assessment for transparency. Splits the VUS category into fVUS (functional unknown) and cVUS (clinical unknown). Introduces "Variant of Interest" (VOI) and "Risk Factor" as clinically meaningful intermediate categories. Endorsed by ESHG working groups. Particularly useful for complex cases where ACMG fails (low penetrance, recessive carriers, non-coding variants, risk alleles).

About

The ABC system is a stepwise variant classification framework proposed by an ad hoc European Society of Human Genetics (ESHG) working group (Houge et al., Eur J Hum Genet 2022;30:150-159). It can classify ANY type of genetic variant — SNVs, indels, CNVs, runs of homozygosity (ROH), enhancer variants, hypomorphic alleles, imprinted alleles, and common risk alleles — which is a major advantage over the ACMG-AMP system that was primarily designed for high-penetrance Mendelian SNVs/indels. Classification proceeds in three steps: (A) Functional grading assesses the variant's biological consequences on gene/protein function, scored 0–5; (B) Clinical grading assesses the genotype–phenotype correlation, scored 0–5; (C) Standard comments linked to the combined A+B class guide reporting decisions. The two-dimensional separation of molecular and clinical evidence makes classification more transparent and reproducible. The system is an add-on or alternative to ACMG-AMP classification and was validated in a 43-laboratory international survey (Houge et al., Eur J Hum Genet 2024).

When to use

Classifying any variant identified on WES, WGS, gene panels, or microarrays — especially variants poorly handled by ACMG (CNVs, ROH, regulatory, hypomorphic)
Evaluating genotype–phenotype correlation independently from molecular predictions
As an add-on to ACMG when a VUS result needs further stratification into fVUS vs. VOI
Deciding whether a variant should be reported in the clinical context — ABC step C standard comments guide reporting
Classifying risk factor variants (e.g., Factor V Leiden, HFE p.Cys282Tyr) that do not fit the binary P/B ACMG paradigm
Multidisciplinary variant review meetings requiring transparent two-dimensional assessment
Re-evaluating historic variants in light of new clinical data or expanded phenotype information

Formula

STEP A — FUNCTIONAL GRADING (variant biology): 0 = fVUS (unknown functional significance) 1 = Normal function (NF) — high-frequency, biologically neutral 2 = Likely normal function (LNF) 3 = Hypothetical functional effect (HFE) — rare, predicted damaging by in silico 4 = Likely functional effect (LFE) / Hypomorphic — likely LoF or partial function 5 = Functional effect (FE) — proven LoF, known GoF, or known dominant-negative STEP B — CLINICAL GRADING (genotype–phenotype fit): 0 = cVUS (unknown clinical significance) 1 = Right type of gene (VOI — variant of interest) 2 = Risk factor (low penetrance, carrier of recessive, or good clinical support) 3 = Pathogenic (penetrance unspecified) 4 = Pathogenic — moderate penetrance (20–40%) 5 = Pathogenic — high penetrance (>40%) COMBINED CLASS (A+B → Final): Class A: A=1–2 (benign/likely benign) → no B grading needed Class B: A=0 (fVUS) or B=0 (cVUS) → VUS Class C: A≥3 + B=1 → Variant of Interest (VOI) Class D: A≥3 + B=2 → Risk Factor Class E: A≥3 + B=3 → Pathogenic (penetrance unspecified) Class E+: A≥3 + B=4 → Pathogenic — moderate penetrance Class F: A≥3 + B=5 → Pathogenic — high penetrance

Pearls & pitfalls

Step A (functional) is INDEPENDENT of clinical phenotype. A truncating variant in BRCA1 always gets A=5 regardless of whether the patient has cancer.
Step B (clinical) evaluates genotype–phenotype FIT. The same variant can get B=5 in a breast cancer patient and B=0 in a patient with unrelated intellectual disability.
The ABC system splits VUS into fVUS (A=0, lack of functional data) and cVUS (B=0, no clinical correlation). This distinction is critical for follow-up prioritization.
A variant classified as A=3 (hypothetical functional effect) + B=1 (right type of gene) = Class C (VOI). This is the most important "rescue" from the ACMG VUS black hole.
Risk Factor (Class D, B=2) captures low-penetrance disease alleles like Factor V Leiden (c.1601G>A), HFE p.Cys282Tyr, and APOE ε4 — variants that ACMG struggles to classify.
In recessive disease, each allele is scored separately in step A. Step B then integrates the combined genotype: a single pathogenic allele = B=2 (carrier/risk), two pathogenic alleles = B=3–5.
The ABC system avoids the paradox where a clearly pathogenic variant (e.g., known LoF in CFTR) is called VUS just because the patient lacks the expected phenotype.
Step C standard comments are linked to each combined class and can be adapted to local or national laboratory policy for reporting thresholds.
Bayesian-calculated ACMG classes and ABC classes mostly overlap for high-penetrance dominant conditions, but diverge for complex, recessive, and low-penetrance scenarios.
In the 43-laboratory international comparison (2024), ABC-based classification showed more consistent reporting decisions than ACMG-based classification for challenging cases.

Next steps

Class A (Benign/Likely Benign): Not reported unless specifically requested. No clinical follow-up needed. Class B (VUS): Report only if laboratory policy requires; no clinical action. Class C (Variant of Interest / VOI): Consider reporting with recommendation for follow-up — segregation analysis, functional studies, or re-review. Class D (Risk Factor): Report with context-specific risk information and management recommendations. Class E (Pathogenic — variable/low penetrance): Report with penetrance information; cascade testing may be offered with appropriate counseling. Class F (Pathogenic — high penetrance): Full clinical reporting; cascade family testing indicated; condition-specific management applies.

Evidence

Houge G et al. Stepwise ABC system for classification of any type of genetic variant. Eur J Hum Genet. 2022;30(2):150-159. PMID:33981013. Validated in: Houge G et al. Comparison of the ABC and ACMG systems for variant classification. Eur J Hum Genet. 2024. PMID:38778080. 43 international laboratories participated. ABC showed more consistent reporting decisions especially for population-frequent disease-associated variants, recessive carriers, and low-penetrance conditions. Commentary: Campeau PM. An all-encompassing variant classification system proposed. Eur J Hum Genet. 2022;30(2):139. Chen JM et al. Expanding ACMG variant classification guidelines into a general framework. Hum Genomics. 2022;16:Article 407.

References

Houge G et al. Stepwise ABC system for classification of any type of genetic variant. Eur J Hum Genet. 2022;30(2):150-159. — link
Houge G et al. Comparison of the ABC and ACMG systems for variant classification. Eur J Hum Genet. 2024. — link
Campeau PM. An all-encompassing variant classification system proposed. Eur J Hum Genet. 2022;30(2):139. — link
Chen JM et al. Expanding ACMG variant classification guidelines into a general framework. Hum Genomics. 2022;16:Article 407. — link
Richards S et al. Standards and guidelines for the interpretation of sequence variants. Genet Med. 2015;17(5):405-424. — link

ABC System for Genetic Variant Classification

Calculator