Variant classifications aren't static. A variant reported as a Variant of Uncertain Significance today might be reclassified as Likely Pathogenic a year from now, or the reverse. For labs and researchers working the ACMG SF v3.3 secondary-findings panel, knowing when and how often that happens matters. ClinVar itself only shows a variant's current status; there's no built-in way to see its history.
I built a small tool to fill that gap.
What it does
The ClinVar Variant History tool tracks classification changes across the 84-gene ACMG SF v3.3 panel, month by month. Search by gene, variant name, or accession, and see the full timeline, which shows where a variant stands today and every classification it has carried across ClinVar's monthly data releases.
The newest addition is a trajectory view. Click into any variant with a classification history and see its path plotted as a step chart. It stays flat while the classification holds and steps up or down at the exact month it changed.
The tool now covers 10 monthly snapshots: January through October 2026. That's 3.0 million records, 341,180 distinct variants, and 9 month-over-month transitions. It refreshes automatically as ClinVar publishes new releases each month.
What 10 months show
Across 341,180 tracked variants, ClinVar's classification label changed 7,869 times between one monthly release and the next, affecting 5,767 distinct variants. Not all of these are clinical reclassifications, as the next section explains.
Examples from the panel: TMEM127 c.149dup moved from Likely Pathogenic to a Pathogenic/Likely Pathogenic consensus between August and September. WT1 c.1447+5G>A held at Pathogenic through June, then shifted to that same Pathogenic/Likely Pathogenic aggregate from July onward. Small movements, but exactly the kind of change that matters if you're relying on a classification from a report filed months or years ago.
Aggregation vs. reinterpretation
Roughly 64% of the 7,869 changes involve a 'Conflicting classifications' label appearing or disappearing. When it appears, that is not someone reclassifying a variant. It is ClinVar's aggregate label shifting because a second submitter has weighed in and disagrees with the first. No existing submitter has changed their view; the label just reflects that two groups now disagree.
March 2026 is a good illustration. That month had 2,259 changes in the panel (the highest single-month total). Of those, 918 were variants moving from Uncertain significance to Conflicting classifications. Typically this means a second submitter classified the variant differently, which tips the aggregate label. The underlying variant wasn't reassessed; ClinVar's aggregation logic changed the label automatically.
If you search a gene and see a large spike in one month, the first question is whether it's driven by Conflicting classifications entries. Often it is.
The RBM20 picture
The clearest outlier once you account for how many variants each gene has is RBM20: 49 changes between definitive classes across 2,425 tracked variants, 48 of them VUS upgrades (44 to Pathogenic, 4 to Likely pathogenic). Per 1,000 tracked variants, that is nearly double the next gene. Rates are shown for genes with 200 or more tracked variants.