ACMG Variant Classification Scale (2D)

Clinical genomics labs don't call a variant "pathogenic" from a single data point — they combine independent evidence criteria under the ACMG/AMP framework, and the combination is what a report is built on. This 2D simulator renders that combination as a physical balance drawn in a draggable perspective view: pathogenic evidence (very strong, strong, moderate) piles weighted discs onto one pan, benign evidence piles onto the other, and the beam tilts by their net weight exactly as Tavtigian's Bayesian point system computes it. A semicircular posterior-probability gauge sweeps across the five reporting tiers — Pathogenic, Likely Pathogenic, VUS, Likely Benign, Benign — while a live stacked bar breaks the total points down criterion-by-criterion, and a draggable prior-probability slider (fixed at 10% in the 3D sibling) shows how the same evidence lands differently in a rarer or more penetrant gene.