Familial hemiplegic migraine: genes and variants
Familial hemiplegic migraine is linked to 1 analyzed protein (ATP1A2). 33 DNA variants are known to cause it; 377 more are uncertain, and 10 of those already look disease-causing on computable evidence.
Last updated 2026-09-30. Research information, not medical advice.
Genes linked to Familial hemiplegic migraine
ATP1A2: Sodium/potassium-transporting ATPase subunit alpha-2
It restores sodium and potassium gradients after activity in astrocytes and other excitable tissues and thereby supports neuronal ion homeostasis. Pathogenic variants are a major cause of familial hemiplegic migraine type 2 and can produce severe episodic neurologic disease.
33 disease-causing and 377 uncertain variants in ATP1A2 are linked to Familial hemiplegic migraine.
Weakly linked (only a few uncertain records): CACNA1A.
Where Familial hemiplegic migraine variants cluster
- ATP1A2 Transmembrane (positions 287–306): 3 of 33 disease-causing changes, 4.6× more than its size predicts.
- ATP1A2 Cytoplasmic (positions 821–840): 3 of 33 disease-causing changes, 4.6× more than its size predicts.
- ATP1A2 Transmembrane (positions 800–820): 3 of 33 disease-causing changes, 4.4× more than its size predicts.
Known disease-causing variants in Familial hemiplegic migraine
| Variant | Position | Protein part | Clinical label |
|---|---|---|---|
| ATP1A2 A606T | 606 | Cytoplasmic | Disease-causing (★★★★) |
| ATP1A2 R548H | 548 | Cytoplasmic | Disease-causing (★★) |
| ATP1A2 G366S | 366 | Cytoplasmic | Disease-causing (★★) |
| ATP1A2 T376M | 376 | Cytoplasmic | Disease-causing (★★) |
| ATP1A2 R383H | 383 | Cytoplasmic | Disease-causing (★★) |
| ATP1A2 G615R | 615 | Cytoplasmic | Disease-causing (★★) |
| ATP1A2 V628M | 628 | Cytoplasmic | Disease-causing (★★) |
| ATP1A2 R834Q | 834 | Cytoplasmic | Disease-causing (★★) |
| ATP1A2 G301R | 301 | Transmembrane | Disease-causing (★★) |
| ATP1A2 T378N | 378 | Cytoplasmic | Disease-causing (★★) |
| ATP1A2 R548C | 548 | Cytoplasmic | Disease-causing (★★) |
| ATP1A2 G855E | 855 | Transmembrane | Disease-causing (★★) |
| ATP1A2 G855R | 855 | Transmembrane | Disease-causing (★★) |
| ATP1A2 R937H | 937 | Cytoplasmic | Disease-causing (★★) |
| ATP1A2 R1002Q | 1002 | Transmembrane | Disease-causing (★★) |
| ATP1A2 G715R | 715 | Cytoplasmic | Disease-causing (★★) |
| ATP1A2 P979L | 979 | Extracellular | Disease-causing (★★) |
| ATP1A2 R202Q | 202 | Cytoplasmic | Disease-causing (★) |
| ATP1A2 R202W | 202 | Cytoplasmic | Disease-causing (★) |
| ATP1A2 G366V | 366 | Cytoplasmic | Disease-causing (★) |
| ATP1A2 I286T | 286 | Cytoplasmic | Disease-causing (★) |
| ATP1A2 T368M | 368 | Cytoplasmic | Disease-causing (★) |
| ATP1A2 R593L | 593 | Cytoplasmic | Disease-causing (★) |
| ATP1A2 M829V | 829 | Cytoplasmic | Disease-causing (★) |
| ATP1A2 R834L | 834 | Cytoplasmic | Disease-causing (★) |
| ATP1A2 G301E | 301 | Transmembrane | Disease-causing (★) |
| ATP1A2 G377V | 377 | Cytoplasmic | Disease-causing (★) |
| ATP1A2 T811R | 811 | Transmembrane | Disease-causing (★) |
| ATP1A2 P786L | 786 | Transmembrane | Disease-causing (★) |
| ATP1A2 D808E | 808 | Transmembrane | Disease-causing (★) |
| ATP1A2 D812H | 812 | Transmembrane | Disease-causing (★) |
| ATP1A2 P331A | 331 | Transmembrane | Disease-causing (★) |
| ATP1A2 A297T | 297 | Transmembrane | Disease-causing |
Uncertain variants in Familial hemiplegic migraine that look disease-causing
| Variant | Position | Protein part | Clinical label | Evidence |
|---|---|---|---|---|
| ATP1A2 R593Q | 593 | Cytoplasmic | Conflicting reports (★) | +7: in a 3D region that tolerates change poorly (4R); R593L at the same position is pathogenic; seen in 6.8e-07 of gnomAD DNA copies; REVEL 0.897 |
| ATP1A2 R383C | 383 | Cytoplasmic | Uncertain (★) | +7: in a 3D region that tolerates change poorly (4R); R383H at the same position is pathogenic; seen in 3.4e-06 of gnomAD DNA copies; REVEL 0.900 |
| ATP1A2 T376R | 376 | Cytoplasmic | Conflicting reports (★) | +6: 3 other pathogenic changes within 3 positions; T376M at the same position is pathogenic; not seen in the gnomAD population database; AlphaMissense 1.00 |
| ATP1A2 T378I | 378 | Cytoplasmic | Conflicting reports (★) | +6: 3 other pathogenic changes within 3 positions; T378N at the same position is pathogenic; not seen in the gnomAD population database; AlphaMissense 1.00 |
| ATP1A2 M829T | 829 | Cytoplasmic | Conflicting reports (★) | +6: in a 3D region that tolerates change poorly (4R); M829V at the same position is pathogenic; not seen in the gnomAD population database; AlphaMissense 1.00 |
| ATP1A2 G366D | 366 | Cytoplasmic | Uncertain (★★) | +6: 3 other pathogenic changes within 3 positions; G366V at the same position is pathogenic; not seen in the gnomAD population database; AlphaMissense 1.00 |
| ATP1A2 R834G | 834 | Cytoplasmic | Uncertain (★) | +6: 2 other pathogenic changes within 3 positions; R834L at the same position is pathogenic; not seen in the gnomAD population database; AlphaMissense 0.99 |
| ATP1A2 R383L | 383 | Cytoplasmic | Uncertain (★★) | +6: in a 3D region that tolerates change poorly (4R); R383H at the same position is pathogenic; not seen in the gnomAD population database; AlphaMissense 0.82 |
| ATP1A2 T811A | 811 | Transmembrane | Uncertain (★) | +6: 3 other pathogenic changes within 3 positions; T811R at the same position is pathogenic; not seen in the gnomAD population database; AlphaMissense 0.96 |
| ATP1A2 D808H | 808 | Transmembrane | Uncertain (★) | +6: 2 other pathogenic changes within 3 positions; D808E at the same position is pathogenic; not seen in the gnomAD population database; AlphaMissense 1.00 |
Which prediction tools work for Familial hemiplegic migraine
How often each tool ranks a disease-causing variant above a harmless one (AUROC × 100).
- MutPred2: 100 out of 100 (learned from overlapping clinical labels, so this is optimistic)
- REVEL: 99 out of 100 (learned from overlapping clinical labels, so this is optimistic)
- AlphaMissense: 99 out of 100
- CADD: 99 out of 100
- CATVariant: 97 out of 100 (learned from overlapping clinical labels, so this is optimistic)
- PolyPhen-2: 93 out of 100 (learned from overlapping clinical labels, so this is optimistic)
- SIFT: 87 out of 100
- phyloP: 82 out of 100
- MetaLR: 73 out of 100 (learned from overlapping clinical labels, so this is optimistic)
Same protein, different disease
- Migraine, familial hemiplegic, 1 is also caused by ATP1A2 variants; they fall mostly in different places as the Familial hemiplegic migraine variants (19 disease-causing).
- Alternating hemiplegia of childhood is also caused by ATP1A2 variants; they fall mostly in different places as the Familial hemiplegic migraine variants (7 disease-causing).
Diseases related to Familial hemiplegic migraine
- Migraine, familial hemiplegic, 1, also linked to ATP1A2
- Alternating hemiplegia of childhood, also linked to ATP1A2
- Fetal akinesia, respiratory insufficiency, microcephaly, polymicrogyria, and dysmorphic facies, also linked to ATP1A2
- Paroxysmal central nervous system disorders, also linked to ATP1A2
Frequently asked questions
Which genes are linked to Familial hemiplegic migraine?
In CATVariant, Familial hemiplegic migraine is linked to 1 analyzed protein: ATP1A2 (Sodium/potassium-transporting ATPase subunit alpha-2).
How many genetic variants are linked to Familial hemiplegic migraine?
433 variants: 33 are classified as disease-causing (pathogenic or likely pathogenic) in ClinVar and 377 are of uncertain significance or have conflicting reports.
Which uncertain variants in Familial hemiplegic migraine look disease-causing?
10 uncertain variants reach the likely-pathogenic range of the ACMG/AMP points scale on computable evidence, for example ATP1A2 R593Q, ATP1A2 R383C, ATP1A2 T376R, ATP1A2 T378I and ATP1A2 M829T. These are leads for expert review, not diagnoses.
Which variant effect predictor works best for Familial hemiplegic migraine?
Among tools not trained on clinical labels, AlphaMissense separates this disease's known disease-causing variants from harmless ones best (AUROC 0.99, based on 17 disease-causing and 10 harmless variants).
About this data
Variant–disease links come from ClinVar, Open Targets and UniProt, pooled from the latest public CATVariant analysis of each human protein. Evidence scores use the ACMG/AMP Bayesian points scale with computable criteria only (position among known disease variants, rarity in gnomAD, calibrated predictors, deep mutational scanning); there is no family or patient data, so they prioritise variants for expert review and never classify them.
Download every variant as CSV · Browse all diseases · Methods · About the Center