Experts in CMT

Prediction: The literature predicts a dominant-negative mechanism for CMT2N: alanyl-tRNA synthetase functions as a homodimer, so a mutant subunit sits in direct contact with its normal counterpart, and the dominant AARS1 substitutions both lower alanyl-tRNA charging and compromise the partner enzyme they assemble with. Supplying additional normal enzyme is not predicted to correct the translational deficit in neurons carrying one of these alleles.

Rationale: Reduced aminoacylation is real but insufficient by itself to produce CMT, and that insufficiency is the supplementation signature of interference rather than of an independent gain: the mutant subunit disrupts the wild-type, so one dominant-negative mechanism accounts for the dominant phenotype. The gain-versus-loss question across the dominant tRNA-synthetase CMTs holds confidence at medium.

Learn more about CMT2N →

Prediction: The literature predicts a dominant-negative mechanism for dHMN-AARS1: alanyl-tRNA synthetase works as a dimer, and the dominant AARS1 alleles yield a subunit that still assembles with its normal partner while failing to charge tRNA-Ala properly. Biallelic AARS1 loss instead causes an early-onset encephalopathy with hypomyelination, and the heterozygous parents in those families have no neuropathy, so one lost copy does not account for this dominant motor phenotype.

Rationale: Halving AARS1 is tolerated, so the dominant alleles must do something a missing copy does not. Humanized yeast co-expression shows mutant AARS1 suppressing growth in the presence of the normal enzyme, and ablating the dimer interface relieves that suppression, which places the defect inside the shared dimer. Whether certain alleles instead acquire a novel binding partner remains argued across the dominant synthetase neuropathies, holding confidence at medium.

Learn more about dHMN-AARS1 →

Mechanistic basis: Complete loss

Prediction: Published evidence points to a biallelic loss-of-function mechanism for CMT-ABHD12: nonsense, frameshift, splice, and missense variants inactivate the serine hydrolase that clears lysophosphatidylserine, and the disease emerges as that signaling lipid accumulates in nervous tissue. The Abhd12 knockout mouse reproduces the lysophosphatidylserine build-up together with progressive auditory and motor decline, and carriers of a single inactive allele remain well, so disease requires both copies to fail.

Rationale: Without functional ABHD12 the damage comes from the substrate, not from a mutant product: lysophosphatidylserine is an immune signaling lipid, and its accumulation provokes the neuroinflammatory and degenerative changes seen in the knockout mouse and in the PHARC phenotype of polyneuropathy, hearing loss, ataxia, retinitis pigmentosa, and cataract. Returning hydrolase activity addresses the primary lesion.

Learn more about CMT-ABHD12 →

Mechanistic basis: Hypomorphic

Prediction: Functional work on the CMTX4 alleles predicts a loss-of-function mechanism: AIFM1 encodes apoptosis-inducing factor, a flavoprotein of the mitochondrial intermembrane space whose partnership with CHCHD4 drives import and folding of the disulfide-relay clients that build the respiratory chain. The reported missense changes destabilize AIF and weaken that interaction, lowering oxidative phosphorylation capacity in peripheral nerve. Males express only the mutant allele, so the hypomorphic protein sets the entire AIF activity of the tissue.

Rationale: The disease alleles act on protein import rather than on apoptotic signaling: destabilized AIF binds CHCHD4 poorly and respiratory subunits fall with it, a shortfall in existing function rather than a new activity. AIF carries several mitochondrial roles, and which one dominates in nerve is unsettled, which is why the mechanism is not graded higher.

Learn more about CMTX4 →

Mechanistic basis: Overactivity

Prediction: The literature predicts a gain-of-function mechanism for CMT-ARHGEF10: the heterozygous Thr109Ile allele, written Thr332Ile on the longer reading frame (c.995C>T), sits at the edge of an autoinhibitory N-terminal region of this RhoA guanine-nucleotide exchange factor. Deleting that region raises exchange activity, and the disease allele does the same, loading more GTP onto RhoA and driving Rho-dependent transcription and ROCK-dependent contraction.

Rationale: Rho-ROCK signaling restrains Schwann cell process extension, so an exchange factor released from its own brake shortens the processes a Schwann cell needs in order to wrap, which is what thin sheaths with uniformly slowed conduction and no axon loss look like. Heterozygous loss of ARHGEF10 is tolerated in population constraint data, so a missing copy is not the lesion. The overactivity rests on a single functional study, holding confidence at medium.

Learn more about CMT-ARHGEF10 →

Prediction: The evidence supports a dominant-negative mechanism for HSN-1D: atlastin-1 drives homotypic endoplasmic reticulum fusion only by dimerizing across apposed tubules, and the heterozygous ATL1 missense alleles yield subunits that still engage those dimers while failing to complete membrane merger, placing the lesion on a shared reaction rather than subtracting from it. Because each failed dimer also occupies a normal subunit, supplying more functional atlastin-1 is not predicted to rebuild the tubular network. Dominant missense alleles at ATL1 also cause SPG3A spastic paraplegia, with no established recessive counterpart.

Rationale: Because fusion needs two atlastin-1 molecules to engage from opposite tubules, a mutant that binds its partner but cannot finish the reaction removes normal protein from productive rounds, an injury different in kind from having half the usual amount. How much simple haploinsufficiency contributes alongside that interference is still argued for ATL1, which is why confidence is held at medium.

Learn more about HSN-1D →

Prediction: The literature predicts a dominant-negative mechanism for HSN-1F: the ATL3 variants Y192C and P338R produce atlastin-3 that still tethers apposed endoplasmic reticulum membranes but arrests before the bilayers merge, and because normal subunits are recruited into the same stalled tether, the mutant withdraws working protein from the fusion cycle instead of simply being absent. A trapped intermediate is evidence of interference rather than a neomorphic activity, though how much fusion capacity survives in patient cells is unresolved.

Rationale: Peripheral sensory axons maintain an endoplasmic reticulum network running their full length, so a tether that holds two membranes together without fusing them leaves that network studded with unresolved junctions. Normal atlastin-3 drawn into a stalled tether is immobilized rather than compensating, so the deficit scales with mutant protein present rather than with protein missing.

Learn more about HSN-1F →

Mechanistic basis: Neomorphic

Prediction: Functional evidence supports a neomorphic gain-of-function mechanism for CMT2DD: the dominant missense substitutions in the alpha1 subunit of the Na/K-ATPase yield a pump that is aberrant rather than absent, and functional dissection of these variants (Clausen et al. 2023) concluded that disease requires expression of the malfunctioning product from one allele and is not haploinsufficiency. An unregulated cation leak through the mutant pump, dissipating the electrochemical gradient that peripheral axons depend on, fits the reported biophysics.

Rationale: A mutant alpha1 pump that leaks cations creates a defect the remaining wild-type copy cannot offset, because the lesion lies in the ion gradient itself rather than in how much transporter is present. That reading follows from the requirement for a malfunctioning product and the explicit exclusion of haploinsufficiency, though a single functional study is thin ground for deciding what the mutant pump actually does.

Learn more about CMT2DD →

Mechanistic basis: Hypomorphic

Prediction: The published evidence supports a loss-of-function mechanism for dSMAX-3: the two ATP7A alleles characterized in X-linked distal hereditary motor neuropathy, p.Thr994Ile and p.Pro1386Ser, retain copper-transport capacity but mislocalize the ATPase, impairing its copper-responsive movement between the trans-Golgi network and the plasma membrane. Hemizygous males have no second ATP7A to compensate, and restored wild-type protein is predicted to rescue a partial, trafficking-level deficit rather than the near-complete deficiency of Menkes disease.

Rationale: The ATP7A allelic series anchors this reading: severe truncating variants produce Menkes disease, milder alleles produce occipital horn syndrome, and these two trafficking-defective missense variants sit at the mild end, with serum copper and ceruloplasmin unremarkable. Whether the motor phenotype follows from that residual deficit alone or from a neuron-specific trafficking requirement is unsettled, which is why confidence stops at medium.

Learn more about dSMAX-3 →

Mechanistic basis: Neomorphic

Prediction: The literature predicts a neomorphic gain-of-function mechanism for CMT2JJ: the recurrent BAG3 p.Pro209Leu substitution sits in the IPV motif that docks HSPB8 and converts the co-chaperone into an aggregation-prone species that pulls HSPB8 and chaperone-assisted selective autophagy clients into insoluble deposits. Truncating BAG3 alleles, which do reduce protein dose, cause dilated cardiomyopathy rather than this childhood-onset neuromuscular presentation, so the CMT2JJ phenotype is a property of the mutant product and added wild-type BAG3 is not predicted to rescue.

Rationale: What drives CMT2JJ is deposit formation rather than reduced BAG3 activity: the p.Pro209Leu product removes HSPB8 and its clients from the soluble pool and stalls proteostasis in muscle and long axons alike. The mechanism rests on a single recurrent allele, and how much of the phenotype comes from the deposits themselves is not resolved, which the medium grade reflects.

Learn more about CMT2JJ →

Mechanistic basis: Overactivity

Prediction: Functional studies predict a gain-of-function mechanism for SMA-LEP-2A: recurrent heterozygous BICD2 substitutions in the N-terminal coiled-coil, p.Ser107Leu foremost among them, weaken the adaptor's closed conformation and increase how readily it recruits and activates dynein-dynactin. Cells expressing the mutant show altered Golgi morphology and disturbed minus-end transport, matching a childhood-onset picture of lower-limb-predominant weakness. The mutant adaptor is overactive in its own right, which separates this mechanism from reduced BICD2 dosage.

Rationale: Enhanced dynein-dynactin recruitment by the mutant adaptor is measurable without any contribution from the normal allele, which is what distinguishes overactivity from interference. Reports that some BICD2 substitutions instead reduce cargo binding keep the mechanism under active argument, and confidence reflects that split rather than any doubt that the disease is dominant.

Learn more about SMA-LEP-2A →

Mechanistic basis: Overactivity

Prediction: Reported evidence predicts overactivity rather than loss in SMA-LEP-2B: the congenital presentation, with arthrogryposis, contractures and respiratory compromise present at birth, arises from BICD2 substitutions that hold the adaptor in an activated conformation and remodel its interactome around Rab6 and the dynein-dynactin motor. Golgi-derived cargo transport is driven rather than blunted. Severity across BICD2 families tracks which substitution is present rather than how much adaptor remains, a distribution that fits activating alleles.

Rationale: Severity tracks the substitution, not the dose: truncating BICD2 alleles are not an established cause of this phenotype, while particular missense substitutions recur at its most severe end. What remains argued is how much of the transport defect comes from excess motor activation and how much from an altered cargo interactome.

Learn more about SMA-LEP-2B →

Mechanistic basis: Neomorphic

Prediction: The literature predicts a toxic gain-of-function mechanism for dHMN-5C: the dominant BSCL2 substitutions N88S and S90L abolish an N-glycosylation site in seipin, and the unglycosylated protein misfolds, is retained in the endoplasmic reticulum, forms inclusions, and triggers an ER stress response that motor neurons tolerate poorly. Biallelic BSCL2 loss instead causes Berardinelli-Seip congenital lipodystrophy, a disease of absent seipin rather than of a retained misfolded one, and adding normal seipin is not predicted to relieve the ER burden.

Rationale: Stripping the glycosylation site turns seipin into an ER-retained aggregate, so the injury comes from what the mutant protein does inside the secretory pathway, not from what the cell is missing. The contrast at this locus is sharp: two null alleles give lipodystrophy with no motor involvement, while one misfolding allele gives distal motor degeneration with no lipodystrophy.

Learn more about dHMN-5C →

Mechanistic basis: Complete loss

Prediction: Genetic evidence assigns CMT-C19orf12 to biallelic loss of function: homozygous truncating and hypomorphic variants, among them the Eastern European founder deletion c.204_214del11, eliminate a small mitochondrial membrane protein tied to lipid and coenzyme A metabolism. The resulting phenotype belongs to the mitochondrial membrane protein-associated neurodegeneration spectrum, with motor neuropathy alongside spasticity, optic atrophy, and brain iron accumulation, and parents carrying one such allele are unaffected.

Rationale: Truncating alleles remove the protein outright, so the phenotype reflects absence rather than a mutant species acting on what remains, and unaffected heterozygous parents in the founder families show that one intact copy supplies enough. That leaves supplementation, not silencing, as the corrective direction the biology points toward.

Learn more about CMT-C19ORF12 →

Prediction: The literature predicts a dominant-negative mechanism for CMT2FF: the recurrent heterozygous CADM3 p.Tyr172Cys substitution introduces an unpaired cysteine that forms an aberrant disulfide bond, leaving the mutant adhesion molecule retained in the endoplasmic reticulum and the CADM3–CADM4 axon-glia adhesion complex disrupted. The variant confers no independent overactive function, so the lesion is the mutant protein's effect on the partner it can no longer present correctly at the membrane rather than any new activity of its own.

Rationale: Absence of any reported CADM3 null or haploinsufficiency CMT argues against halved dosage as the lesion, while the recurrent allele is characterized as a mistrafficked product that pulls its binding partner out of service at the axon-glia interface. Characterization rests on one recurrent variant and a single group's functional work, which caps confidence at medium.

Learn more about CMT2FF →

Mechanistic basis: Hypomorphic

Prediction: One recurrent chaperonin substitution carries this subtype, and the literature predicts recessive loss of function: the homozygous CCT5 His147Arg change alters a subunit of the CCT/TRiC folding chamber, reducing the complex's capacity to fold actin, tubulin, and other obligate clients in long sensory and corticospinal axons. Two defective copies are required for the mutilating sensory loss with spastic paraplegia, and restoring folding-competent CCT5 is predicted to rescue.

Rationale: A subunit of a hetero-oligomeric ring is the classic setting for interference, so a dominant-negative effect is worth entertaining here, but heterozygous relatives are unaffected and the phenotype requires homozygosity. How much folding activity a His147Arg-containing ring retains has not been fully resolved, and that gap sets the grade at medium.

Learn more about HSN w/SPG →

Mechanistic basis: Haploinsufficiency

Prediction: The literature predicts a loss-of-function mechanism for CMT-CFAP276: Sun et al. (2019) reported heterozygous CFAP276 (C1orf194) variants p.Lys28Ile and p.Ile122Asn, and p.Lys28Ile lowers steady-state protein. Heterozygous null mice develop a dominant-intermediate CMT with reduced conduction velocity, demyelination and shortened internodes, and AAV9 delivery of the human coding sequence corrects motor and nerve measures in nulls, placing the lesion in how much functional protein the nerve retains.

Rationale: One inactivated copy suffices in the mouse and adding the gene back repairs the nerve, the pairing a dosage-sensitive loss predicts. The p.Ile122Asn product is unusually stable and forms cytoplasmic aggregates, but it has not been shown to be toxic or to acquire an activity of its own, and a knock-in of that allele reproduces the phenotype the null already produces. Confidence holds at medium while the aggregating allele stays unexplained.

Learn more about CMT-CFAP276 →

Mechanistic basis: Neomorphic

Prediction: The evidence predicts a toxic gain-of-function mechanism for CMT-CHCHD10: heterozygous S59L and R15S alleles make CHCHD10 aggregation-prone inside the mitochondrial intermembrane space, where the misfolded protein accumulates, sequesters its paralog CHCHD2, and provokes the mitochondrial integrated stress response. Disease therefore tracks a species the mutant acquires, not a shortfall in normal CHCHD10 activity, though part of the field reads the CHCHD2 co-sequestration as interference with wild-type protein instead.

Rationale: Aggregation is what these alleles add, not what they subtract, and the healthy Chchd10-knockout mouse is why haploinsufficiency does not explain the dominant phenotype: complete absence of the protein is tolerated in that model. Whether trapping CHCHD2 alongside the mutant also amounts to dominant-negative interference is unresolved, and the medium grade follows from that.

Learn more about CMT-CHCHD10 →

Mechanistic basis: Complete loss

Prediction: The literature predicts a biallelic loss-of-function mechanism for CMT-CNTNAP1: nonsense, frameshift, and missense variants abolish or destabilize Caspr, the axonal neurexin-family protein that pairs with contactin-1 and glial neurofascin-155 to build the paranodal septate-like junction. Cntnap1-null mice lose the transverse bands of that junction and let juxtaparanodal potassium channels spread into the paranode, degrading saltatory conduction, the same failure seen in patients carrying two defective copies.

Rationale: Caspr loss dismantles the axo-glial seal at the paranode, and the conduction findings follow from missing adhesion rather than from a mutant protein sitting in the junction. Missense alleles that cannot engage contactin-1 never reach the axon surface, which makes them functionally null and leaves nothing at the paranode to interfere with.

Learn more about CMT-CNTNAP1 →

Mechanistic basis: Hypomorphic

Prediction: Functional studies support a biallelic loss-of-function mechanism for CMT-COA7: homozygous and compound-heterozygous variants deplete a Sel1-repeat protein of the mitochondrial intermembrane space that acts as a respiratory chain assembly factor, and patient fibroblasts show reduced complex I and complex IV activity. Re-expressing wild-type COA7 in those cells returns respiratory chain assembly toward normal, the clearest available evidence that these alleles subtract function rather than add a harmful one.

Rationale: Patient cells lose cytochrome c oxidase activity because the assembly factor is missing, not because an altered version obstructs assembly, and the complementation result settles that point directly. The clinical picture of ataxia with sensorimotor involvement follows the same bioenergetic shortfall in cerebellum and in long axons.

Learn more about CMT-COA7 →

Mechanistic basis: Hypomorphic

Prediction: Hypomorphic COQ7 alleles are predicted by the literature to act through biallelic loss of function in dHMN-COQ7: COQ7 hydroxylates demethoxyubiquinone in the closing steps of coenzyme Q10 synthesis, so when both copies are impaired DMQ10 accumulates while CoQ10 falls, starving motor axons of mitochondrial electron transport. Patient cells treated with 2,4-dihydroxybenzoate, which bypasses the missing hydroxylation step, recover CoQ10, and that biochemical correction is a direct test of the deficit.

Rationale: A missing enzymatic step, visible as DMQ10 buildup alongside reduced CoQ10, defines this subtype, and metabolite bypass corrects the deficiency in patient fibroblasts. Correction by a downstream intermediate is possible only if the pathology is insufficient product, not a harmful mutant enzyme.

Learn more about dHMN-COQ7 →

Mechanistic basis: Complete loss

Prediction: Assays in patient fibroblasts support a biallelic loss-of-function mechanism for CMTRID: COX6A1 encodes a nuclear-encoded structural subunit of cytochrome c oxidase, and the recessive splice-region and loss variants remove functional subunit from both copies, leaving complex IV under-assembled with measurably reduced activity. The deficit is quantitative, scaling with how much intact subunit survives, and one preserved allele is enough to keep carriers healthy.

Rationale: A structural subunit missing from the holoenzyme lowers the amount of assembled cytochrome c oxidase rather than producing an aberrant one, and fibroblast measurements track that shortfall. Why peripheral nerve bears the burden while other high-demand tissues are spared has not been explained, and with few reported families the mechanism is graded medium.

Learn more about CMTRID →

Prediction: The literature predicts a dominant-negative mechanism for CMT-CRYAB: alphaB-crystallin functions only as a large oligomer and exchanges subunits with alphaA-crystallin, so the heterozygous p.Arg120Gly allele reported by Cortese et al. (2023) yields a subunit that enters those complexes and degrades the chaperone capacity of the whole assembly. Complete CRYAB nulls instead cause a recessive infantile myofibrillar myopathy whose heterozygous carriers are well, so a halved dose does not produce this dominant disease.

Rationale: Mutant alphaB-crystallin does not stay apart from the normal protein: it assembles into an irregular high-molecular-weight complex, and co-expressing other small heat-shock chaperones relieves the aggregation, which points to capacity lost inside a shared oligomer rather than a toxin acting alone. A 2025 structural report instead reads the mutant as a constitutively activated chaperone, and confidence stays low while that reading is open.

Learn more about CMT-CRYAB →

Mechanistic basis: Hypomorphic

Prediction: The literature predicts a biallelic loss-of-function mechanism for CMT-CTDP1: the founder intronic retrotransposon insertion (IVS6+389C>T) causes partial mis-splicing of CTDP1 that lowers functional FCP1 phosphatase across both alleles, a hypomorphic reduction rather than a toxic product. Homozygosity for this Roma founder allele accounts for essentially all reported congenital cataracts facial dysmorphism neuropathy families, and single-allele carriers are unaffected. Restored wild-type is predicted to rescue, consistent with recessive loss of function.

Rationale: Mis-splicing is partial, and that is the point: FCP1, the phosphatase that resets the RNA polymerase II carboxy-terminal domain between transcription cycles, is essential, so complete absence is not compatible with development. What the founder insertion leaves behind is a dosage floor low enough to cause disease and high enough to permit survival, not an interfering product.

Learn more about CMT-CTDP1 →

Mechanistic basis: Hypomorphic

Prediction: Evidence from the reported genotypes supports biallelic loss of function in CMT-DARS2: a leaky intron-2 splice variant causing partial skipping of exon 3 sits in trans with a severe or null allele in nearly every patient, leaving reduced but non-zero mitochondrial aspartyl-tRNA synthetase and impairing mitochondrial translation. Two null alleles have not been observed, and the residual activity the leaky allele preserves appears to be what makes the genotype survivable.

Rationale: The leaky splice allele carries the phenotype, and its tissue-variable exon 3 skipping explains the range from leukoencephalopathy with brainstem and spinal cord involvement to a predominantly peripheral presentation. A hypomorphic enzyme deficiency of this shape answers to added functional synthetase, and nothing in the genotypes indicates a mutant enzyme acting against its wild-type partner.

Learn more about CMT-DARS2 →

Mechanistic basis: Unresolved

Prediction: Evidence for GAN-2 comes from a single family carrying dominant DCAF8 missense variants, and the literature stops short of a mechanism. DCAF8 serves as a substrate receptor for the CRL4 ubiquitin-ligase complex, so a halved supply of receptor and an inert receptor that occupies the complex without delivering substrate are both plausible readings of the same phenotype, and no functional work separates them. Biallelic gigaxonin loss causes the recessive giant axonal neuropathy GAN-1.

Rationale: A substrate receptor can fail either by being absent or by docking into the ligase and blocking productive ubiquitination, and with one reported pedigree and no biochemistry on the mutant DCAF8, neither reading can be excluded. That is a different genetic situation from GAN-1, where gigaxonin loss and intermediate-filament disorganization are established.

Learn more about GAN-2 →

Prediction: The evidence supports a dominant-negative mechanism for dHMN-7B: the CAP-Gly substitution p.Gly59Ser destabilizes the microtubule-binding domain of p150glued, and the misfolded subunit both assembles into dynactin and accumulates in cytoplasmic aggregates that draw normal p150glued in with it, so retrograde transport degrades further than one working allele would predict. While aggregates keep sequestering normal subunit, added wild-type p150glued is not predicted to rescue transport. Other dominant DCTN1 alleles cause Perry syndrome, and no recessive DCTN1 disease supports a haploinsufficiency route to this phenotype.

Rationale: Aggregated Gly59Ser p150glued withdraws normal subunit from circulation, so the cell loses more dynactin activity than one mutant allele accounts for, and retrograde transport fails in the motor neurons that declare themselves here as vocal cord paralysis and hand wasting. Whether those aggregates additionally act as a toxic species, independent of what they sequester, is unsettled.

Learn more about dHMN-7B →

Mechanistic basis: Unresolved

Prediction: A mechanism for CMT-DGAT2 is not established in the literature: a single autosomal dominant family with early-onset CMT (Hong et al. 2016) carries a variant in DGAT2, the endoplasmic reticulum enzyme that performs the committed final step of triacylglycerol synthesis and channels diacylglycerol into lipid droplets. Whether the disease follows from reduced triglyceride output or from accumulation of diacylglycerol, itself a bioactive signaling lipid, has not been tested, and no dosage or rescue experiments have been reported.

Rationale: A dominant perturbation of DGAT2-mediated lipid handling is what the family demonstrates, and the two readings of it diverge sharply: a shortfall in triacylglycerol synthesis is a quantitative lesion, while diacylglycerol accumulation would be toxic in a way that need not scale with enzyme level. One family and no enzyme kinetics cannot choose between them.

Learn more about CMT-DGAT2 →

Mechanistic basis: Haploinsufficiency

Prediction: DHTKD1 encodes the E1 component of the 2-oxoadipate dehydrogenase complex in the lysine degradation pathway. For CMT2Q the literature predicts haploinsufficiency as the route to loss of function: the heterozygous nonsense allele p.Y486* reported by Xu et al. (2012) lowered DHTKD1 transcript and protein in patient cells, with reduced ATP production and elevated reactive oxygen species. Because no truncated product is retained, added wild-type protein is predicted to rescue.

Rationale: Nonsense-mediated decay of the p.Y486* transcript removes the message rather than delivering a truncated enzyme into the complex, so the deficit reads as dosage. Two things temper the evidence: only one dominant family has been described, and complete biallelic DHTKD1 loss causes a separate recessive disease, 2-aminoadipic 2-oxoadipic aciduria (AMOXAD), leaving it unsettled whether one lost copy suffices for dominant CMT. Confidence sits at medium.

Learn more about CMT2Q →

Mechanistic basis: Unresolved

Prediction: The literature does not resolve a mechanism for CMT-DHX9: heterozygous missense variants in the helicase core segregate with dominant CMT, and DHX9 unwinds DNA:RNA hybrids, G-quadruplexes and structured RNA, so a catalytically impaired subunit could lower net helicase output or instead remain bound to those substrates and block wild-type access to them. The 2023 cohort reports no unwinding or R-loop measurements on the patient alleles.

Rationale: The CMT-associated alleles are missense rather than the loss-of-function variants found elsewhere in DHX9, which weakens a straightforward dosage account without arguing for any one alternative. Telling an inert mutant that occupies R-loop substrates from one with altered unwinding kinetics requires biochemistry nobody has performed.

Learn more about CMT-DHX9 →

Mechanistic basis: Complete loss

Prediction: The literature predicts a biallelic loss-of-function mechanism for dSMA-5: DNAJB2 encodes the HSJ1 Hsp40 co-chaperone, which uses its ubiquitin-interacting motifs to capture ubiquitylated substrates and hand them to Hsp70 for refolding or proteasomal disposal. The recessive splice, truncating, and destabilizing missense alleles subtract that activity from both copies, and add-back of wild-type DNAJB2 is predicted to restore substrate handling.

Rationale: Clearance capacity scales with how much co-chaperone is available, so homozygous splice and truncating alleles leave motor neurons unable to process ubiquitylated clients while carriers with one working copy have enough. Nothing about the mutant products suggests they occupy Hsp70 or block its remaining partners.

Learn more about dSMA-5 →

Prediction: Published functional work supports a dominant-negative mechanism for CMT2M: dynamin-2 severs membrane necks during endocytosis and intracellular traffic only when many copies assemble into a helical collar, and the CMT-associated PH-domain substitutions, which weaken phosphoinositide binding, are built into those collars alongside normal subunits, stalling constriction that intact subunits alone would complete. Added wild-type protein is diluted into the same mixed polymers instead of replacing them.

Rationale: One defective subunit is enough to stall a collar whose constriction depends on the whole polymer acting in concert, and no CMT has been attributed to simple DNM2 dosage reduction. What stays contested is whether these alleles also gain activity, since the enhanced-assembly evidence comes from centronuclear-myopathy substitutions rather than the CMT set, leaving the mechanism unsettled at medium confidence.

Learn more about CMT2M →

Mechanistic basis: Overactivity

Prediction: The literature predicts an overactivity gain-of-function mechanism for CMTDIB: PH-domain changes in DNM2 yield a dynamin-2 that self-assembles more readily and carries elevated assembly-stimulated GTPase activity, so endocytic membrane fission runs faster and less selectively than the cell can regulate. That excess is intrinsic to the mutant and does not require the wild-type partner, and a single null DNM2 allele does not reproduce the phenotype. Whether the dominant effect is purely overactivity or also involves mixed dynamin polymers is argued, so confidence is medium.

Rationale: Faster and more avid self-assembly means the mutant dynamin fissions membranes on its own account, so the trafficking defect follows from surplus activity rather than from any block imposed on normal dynamin-2. Hyperassembling DNM2 alleles produce dominant centronuclear myopathy by the same overactive route, reinforcing that reading. How much of the effect persists once mutant and wild-type co-assemble is the open question behind the medium grade.

Learn more about CMTDIB →

Mechanistic basis: Mixed

Prediction: For HSN-1E the literature predicts one dominant allele class carrying two mechanisms at once. The DNMT1 RFTS-domain variants misfold and are cleared prematurely, lowering maintenance methyltransferase activity and leaving a globally hypomethylated methylome, a partial loss. The same variants mislocalize and aggregate, adding a toxic species on top of that deficit, and heterozygous DNMT1 nulls do not reproduce the disease. Both arms together support a mixed mechanism.

Rationale: A missing DNMT1 copy does not cause the disease, so lost maintenance methylation cannot be the whole account; the mislocalized, aggregating mutant supplies the dominant component that haploinsufficiency lacks. How much of the phenotype that species drives on its own is still unsettled.

Learn more about HSN-1E →

Mechanistic basis: Haploinsufficiency

Prediction: Loss of function is the mechanism the literature assigns to CMT-DRP2: the reported X-linked nonsense variant eliminates dystrophin-related protein 2, which anchors periaxin to the dystroglycan complex in myelinating Schwann cells and organizes the appositions that partition Cajal bands. Drp2-null mice lose those appositions and develop demyelination. In hemizygous males the abolished copy is the only source of the protein, leaving the Schwann cell membrane scaffold with nothing to fall back on.

Rationale: Absent DRP2 leaves periaxin without its membrane partner and the Cajal band architecture unpartitioned, a structural shortfall a functional copy would fill rather than a mutant scaffold competing for a place in the complex. Human evidence rests on very few families, with the mouse carrying most of the mechanistic weight, which is why confidence is medium rather than high.

Learn more about CMT-DRP2 →

Mechanistic basis: Complete loss

Prediction: Reported families and the dystonia musculorum mouse point to biallelic loss of function in CMT-DST: compound-heterozygous and null DST variants remove the neuronal isoform of dystonin, the plakin-family linker that ties neurofilaments and microtubules to the actin cytoskeleton in sensory neurons, while sparing the epithelial isoform (Motley et al. 2020). Heterozygous carriers are unaffected, and the spontaneous null mouse degenerates its sensory neurons.

Rationale: Isoform-specific truncation is the discriminating detail: variants confined to the neuronal transcript leave epidermal dystonin intact, which is why these patients have sensory loss without skin fragility. Confidence is medium because the human series is small and DST is a large, variant-tolerant gene in which assigning pathogenicity to a given allele pair is rarely straightforward.

Learn more about CMT-DST →

Mechanistic basis: Complete loss

Prediction: Mouse and human genetics converge in HSAN-6, where the literature predicts biallelic loss of function: homozygous truncating DST alleles remove dystonin, the cytoskeletal linker that couples actin, microtubules, and intermediate filaments in sensory neurons, and Dst-null dystonia musculorum mice reproduce the sensory ganglion degeneration seen in patients. One intact allele is sufficient in carriers, so the human disease tracks the mouse null.

Rationale: The mouse carries most of the weight here: complete Dst loss reproduces the sensory neurodegeneration, tying the human truncating alleles to absent dystonin rather than to a mutant isoform interfering with cytoskeletal cross-linking. Human confirmation still rests on a small number of families with a narrow allele spectrum, and that limitation is what stops the grade short of high.

Learn more about HSAN-6 →

Prediction: Mouse and human evidence supports a dominant-negative mechanism for CMT2O: cytoplasmic dynein hauls cargo as a two-headed complex built on a pair of DYNC1H1 heavy chains, and a heterozygous missense chain is assembled into that motor next to its normal partner, slowing retrograde transport that an all-normal motor completes. Heterozygous DYNC1H1 loss produces a neurodevelopmental phenotype rather than this CMT, while the Loa mouse, carrying a single heavy-chain point mutation, degenerates.

Rationale: Every dynein motor carries two heavy chains, so in a heterozygote most motors contain a mutant one and the defect travels with the cargo rather than being averaged out across the cell. How much altered motor behavior contributes alongside that interference is still open, and the residual uncertainty in this mechanism sits there.

Learn more about CMT2O →

Prediction: The evidence predicts a dominant-negative mechanism for SMA-LEP-1: the tail-domain and motor-domain missense variants behind this dominant, lower-limb-predominant spinal muscular atrophy leave the heavy chain able to homodimerize, so most cytoplasmic dynein-1 motors in a heterozygote carry a mutant chain and move cargo poorly in the motor neurons supplying the legs. Extra wild-type heavy chain is not predicted to displace a mutant chain already assembled into a dimer. DYNC1H1 haploinsufficiency produces a neurodevelopmental phenotype instead, and whether the motor deficit reflects interference or reduced effective dynein is unsettled, so confidence sits at medium.

Rationale: Weakness here is proximal, largely static, and predominant in the legs, a distribution pointing to a transport defect in one motor pool rather than a length-dependent process. Two heavy chains build each dynein-1 motor, so three quarters of motors in a heterozygote contain a mutant one. The Loa mouse, carrying a single Dync1h1 missense change alongside an intact wild-type allele, develops the same dominant motor degeneration.

Learn more about SMA-LEP-1 →

Prediction: The literature predicts a dominant-negative mechanism for CMT1D: heterozygous EGR2 mutations, mostly in the zinc-finger DNA-binding domain and the R1/NAB-interaction domain, produce a mutant transcription factor that fails to activate myelination genes yet interferes with the wild-type protein. Zinc-finger mutants reduce DNA binding and act dominant-negatively, while R1-domain mutants disrupt the NAB corepressor interaction. Functional studies showing impaired transactivation with dominant suppression of wild-type EGR2 activity indicate that added wild-type protein is not predicted to rescue, arguing against simple haploinsufficiency.

Rationale: Heterozygous zinc-finger EGR2 mutants lose DNA binding yet disrupt wild-type Krox20 transactivation, which points to a dominant-negative effect rather than simple loss of function. The R1-domain/NAB-disruption alleles carry a gain-of-function-like deregulation that keeps the overall mechanism partly debated, holding confidence at medium.

Learn more about CMT1D →

Mechanistic basis: Hypomorphic

Prediction: EGR2/Krox20 is the transcription factor that drives the Schwann cell myelination program, and the literature predicts that CMT4E follows from lowering its output on both alleles: the recessive substitutions, notably p.Ile268Asn in the R1 domain, reduce EGR2 transactivation, and homozygotes present with congenital hypomyelination while heterozygous parents are normal. The dominant zinc-finger EGR2 variants take the opposite route, yielding a DNA-binding-defective protein that disrupts wild-type Krox20 in CMT1D.

Rationale: One gene, two mechanisms, sorted by allele class: an R1-domain recessive change lowers transcriptional output and needs a second defective copy to be felt, while zinc-finger changes are dominant because the mutant protein acts on the normal one. Whether p.Ile268Asn is purely hypomorphic or also deranges NAB corepressor binding is unresolved, and the medium grade reflects that.

Learn more about CMT4E →

Mechanistic basis: Hypomorphic

Prediction: The literature predicts a biallelic loss-of-function mechanism for HSAN-3 (familial dysautonomia): nearly every patient is homozygous for the Ashkenazi founder allele ELP1 c.2204+6T>C, an intron 20 splice variant that is leaky and tissue-specific, so exon 20 is skipped most heavily in nervous tissue while other tissues retain substantial correctly spliced transcript. The resulting shortfall of Elongator subunit ELP1 in sensory and autonomic neurons, not a toxic truncated product, produces the disease.

Rationale: Tissue-specific skipping explains the selective neuronal vulnerability of an otherwise ubiquitous complex, and it is why the founder allele is survivable where a complete null would not be. Splicing modulators that raise exon 20 inclusion restore full-length ELP1 and are predicted to rescue, the therapeutic reading of a hypomorphic, quantity-limited defect.

Learn more about HSAN-3 →

Mechanistic basis: Unresolved

Prediction: The literature stops short of a mechanism for CMT1H: the recurrent dominant missense allele alters fibulin-5, a secreted matrix glycoprotein that tethers lysyl oxidase-like enzymes to tropoelastin and organizes elastic fiber assembly. Since fibulin-5 multimerizes outside the cell, a mutant that is still secreted could disrupt assembly directly, whereas one retained inside would only reduce the protein available, and no study has determined which the CMT1H allele does.

Rationale: Biallelic FBLN5 loss of function causes autosomal recessive cutis laxa rather than CMT, so the dominant nerve phenotype is not a simple reading of reduced fibulin-5. Whether the missense product reaches the matrix and interferes there, or never arrives at all, is the untested step on which the mechanism turns.

Learn more about CMT1H →

Mechanistic basis: Haploinsufficiency

Prediction: Dosage, not interference, is what the literature predicts for dHMN-2D: the recurrent heterozygous p.Cys206Arg substitution (Sumner et al.) weakens FBXO38 in its role as the SCF substrate-recognition subunit that coactivates KLF7-dependent transcription of axonal growth genes, and motor neurons are sensitive to how much of that transcriptional program reaches them. One reduced copy is therefore sufficient to produce the dominant phenotype, and the defect is lowered output rather than a toxic product.

Rationale: Reduced KLF7 coactivation is the measured defect, and a quantitative shortfall in a dosage-sensitive transcriptional program accounts for dominant transmission without invoking a harmful protein. A single recurrent missense in a substrate-recognition subunit could still disrupt assembly of the SCF complex, and that unexcluded alternative is why confidence sits at medium.

Learn more about dHMN-2D →

Mechanistic basis: Complete loss

Prediction: The literature predicts a biallelic loss-of-function mechanism for CMT4H: recessive FGD4 nonsense, frameshift, splice, and missense variants hitting the FYVE, DH, and PH domains take away frabin's ability to load GTP onto Cdc42 at PI3P-marked membranes, so the Schwann cell loses the cytoskeletal remodeling that shapes its sheath. Heterozygous parents are clinically normal, and mouse work deleting Fgd4 in Schwann cells places the defect inside the myelinating cell itself.

Rationale: Frabin sits upstream of Cdc42, so removing it silences a signaling step instead of adding a new activity, and the domain-targeted missense variants behave like the truncations in converging on absent exchange activity. Outfolding pathology shared with the myotubularin subtypes puts both on phosphoinositide-directed membrane remodeling.

Learn more about CMT4H →

Mechanistic basis: Hypomorphic

Prediction: FIG4 is unusual in how its alleles assort: CMT4J patients almost always carry the hypomorphic p.Ile41Thr substitution, which destabilizes FIG4 within the PIKFYVE and VAC14 complex, in trans with a null allele, so PI(3,5)P2 regulatory activity falls without reaching zero. The literature predicts that this residual amount sets severity, a reading supported by the pale tremor mouse and by complementation in Fig4-null cells. Two null alleles together instead produce Yunis-Varon syndrome, a severe multisystem developmental disorder.

Rationale: Severity that tracks with how much FIG4 protein survives is the signature of a quantitative deficit, and the p.Ile41Thr product is destabilized rather than newly active. Heterozygous FIG4 variants segregating with ALS11 show what one altered allele does on its own, which is not to produce CMT4J.

Learn more about CMT4J →

Mechanistic basis: Hypomorphic

Prediction: Published evidence predicts a biallelic loss-of-function mechanism for HSN-FLVCR1: homozygous and compound-heterozygous variants strip the FLVCR1 transporter of the activity it needs for both heme export and choline import, leaving sensory neurons without heme detoxification or an adequate phospholipid precursor supply. Mice lacking Flvcr1a lose dorsal root ganglion sensory neurons, and the same allelic series also produces posterior column ataxia with retinitis pigmentosa.

Rationale: Transport capacity falls only when both alleles are hit, and heterozygous relatives across the reported pedigrees are unaffected, so the lesion is a quantitative shortfall of FLVCR1 activity in dorsal root ganglion neurons. Restoring wild-type FLVCR1 is therefore predicted to rescue.

Learn more about HSN-FLVCR1 →

Mechanistic basis: Complete loss

Prediction: Gene-replacement data and the recessive genetics together predict loss of function in GAN-1: gigaxonin is the substrate adaptor that delivers intermediate filaments to a Cullin3 ubiquitin ligase for degradation, and the GAN mutations abolish or destabilize it, so neurofilaments accumulate and distend axons into the swellings that name giant axonal neuropathy. Both copies must be affected and carrier parents are well. Reintroducing wild-type gigaxonin restores filament turnover, the premise of the intrathecal gene-transfer program.

Rationale: Intermediate filaments have no alternative route to degradation once the adaptor is gone, and the packed giant axons are the anatomical record of that failure. Clearance resumes when gigaxonin is reintroduced into patient cells, which is the strongest argument that the mutant alleles subtract a function rather than add a pathological one.

Learn more about GAN-1 →

Mechanistic basis: Neomorphic

Prediction: The literature predicts a neomorphic gain-of-function mechanism for CMT2D: heterozygous GARS1 missense alleles open the enzyme's conformation and expose surfaces the wild-type synthetase keeps buried, creating interactions it never makes, including inappropriate binding to Nrp1 and to Trk receptors and mislocalization within peripheral axons. Heterozygous Gars1 null mice do not develop CMT, so the dominant phenotype is not a matter of reduced enzyme dose. Whether the toxic species acts chiefly through aberrant receptor engagement or through sequestration of tRNA-Gly remains open, and confidence is graded medium.

Rationale: These alleles do not lose activity so much as acquire it, opening a conformation that exposes new surfaces and engaging Nrp1 and Trk receptors on motor axons. Severity across the CMT2D alleles tracks that aberrant surface rather than residual aminoacylation, and overexpressed wild-type GARS1 does not rescue, which separates a toxic gain both from reduced enzyme dose and from interference with the normal subunit.

Learn more about CMT2D →

Mechanistic basis: Neomorphic

Prediction: Published evidence predicts a neomorphic gain-of-function mechanism for dHMN-5A: dominant GARS1 substitutions destabilize the enzyme into an open conformation that exposes a surface absent from wild-type glycyl-tRNA synthetase, permitting aberrant binding to Nrp1 with antagonism of VEGF/Nrp1 signaling, aberrant engagement of Trk receptors, and mislocalization within motor axons. Null and loss-of-function GARS1 alleles do not produce the dHMN phenotype, and severity does not track residual aminoacylation, which excludes haploinsufficiency. How much a partial loss of aminoacylation still contributes is argued, and confidence is medium.

Rationale: The open conformation is the whole lesion: it creates a binding surface, and the Nrp1 and Trk interactions follow from it independently of what the enzyme still does to tRNA. Carriers of nulls are unaffected and the most severe alleles are not the least catalytically competent, so extra normal enzyme reaches neither event. The residual-catalysis question is what holds the grade at medium.

Learn more about dHMN-5A →

Mechanistic basis: Unresolved

Prediction: The GBF1 data do not narrow CMT2GG to one mechanism: ultra-rare dominant and de novo GBF1 missense variants cause CMT with Golgi fragmentation, and GBF1 is the exchange factor that loads ARF1 with GTP at the cis-Golgi to recruit COPI coats. Fragmentation reports a shortfall of ARF1-GTP without reporting its cause, which leaves reduced exchange output and a mutant GEF that engages ARF1 unproductively both standing.

Rationale: Golgi fragmentation is the shared endpoint of every route to insufficient ARF1 activation, so observing it constrains nothing about how the mutant arrives there. Nucleotide-exchange assays on the variant proteins would tell reduced output from active interference, and none have been reported.

Learn more about CMT2GG →

Mechanistic basis: Complete loss

Prediction: GDAP1 sits in the outer mitochondrial membrane and promotes mitochondrial fission, and its loss leaves an elongated network that traffics poorly along long axons. The literature supports a biallelic loss-of-function mechanism for CMT2B3, driven by null or hypomorphic alleles on both chromosomes, with the Gdap1 knockout mouse reproducing the peripheral phenotype. Reintroduced GDAP1 is predicted to restore fission.

Rationale: Both GDAP1 copies must fail before fission falls far enough to injure nerve, and the knockout mouse demonstrates that removing the protein outright is sufficient to produce the phenotype. The dominant GDAP1 alleles, reported under CMT2K alongside recessive ones, act by a different route, disrupting wild-type GDAP1 within the outer membrane. Allele class, not the gene, sets the mechanism.

Learn more about CMT2B3 →

Mechanistic basis: Mixed

Prediction: Evidence in CMT2K divides along inheritance. Biallelic GDAP1 alleles are loss-of-function or null, supported by the Gdap1 knockout mouse in which restored wild-type protein corrects the deficit, whereas the dominant GST-domain alleles yield a mutant GDAP1 that disrupts GDAP1-dependent mitochondrial fission in the presence of a normal copy. Since one subtype sustains a recessive null mechanism and a dominant interfering one, the literature supports a mixed mechanism.

Rationale: The two allele classes converge on mitochondrial fission by different routes, total absence of GDAP1 from both chromosomes on one side and a single GST-domain mutant that disrupts the wild-type subunit beside it on the other. The dominant route is contested, since halved GDAP1 would account for much of the milder phenotype on its own, and that dispute limits confidence to medium.

Learn more about CMT2K →

Mechanistic basis: Complete loss

Prediction: The literature predicts a biallelic loss-of-function mechanism for CMT4A: recessive GDAP1 nonsense, frameshift, and missense alleles strip this outer mitochondrial membrane protein of its role in mitochondrial fission, and Gdap1-null mice reproduce the peripheral nerve pathology. Two defective copies are required, carriers are clinically unaffected, and mitochondrial dynamics are predicted to recover if GDAP1 is supplied. Heterozygous GDAP1 missense alleles that act on the wild-type protein instead produce the dominant subtype CMT2K.

Rationale: Loss of GDAP1 leaves Schwann cells and axons with a mitochondrial network they cannot remodel, and the null mouse phenocopies the human disease. The early severe course, with vocal cord paresis and diaphragmatic weakness, tracks with absence of the protein rather than with a product that interferes with its normal counterpart, which is what dominant GDAP1 alleles do in CMT2K.

Learn more about CMT4A →

Mechanistic basis: Complete loss

Prediction: Evidence from the recessive GDAP1 families predicts a loss mechanism requiring both copies in CMTRIA: GDAP1 sits in the outer mitochondrial membrane, where its glutathione-S-transferase-like domains support mitochondrial fission and the cellular antioxidant response, and these alleles strip that activity from the protein. GDAP1 is required in Schwann cells and in the neuron alike, which is why conduction in these patients settles between the demyelinating and axonal ranges instead of within either. Heterozygous parents are clinically normal.

Rationale: The lost protein is needed on both sides of the myelinated unit, so a single depletion is read out simultaneously on myelin and on the axon, giving the intermediate electrophysiology that defines this presentation. Truncating and severe missense alleles behave alike when homozygous, while the dominant GDAP1 missense alleles that interfere with the wild-type protein produce a separate dominant subtype, CMT2K.

Learn more about CMTRIA →

Mechanistic basis: Haploinsufficiency

Prediction: The literature predicts a loss-of-function mechanism for CMTX1: connexin 32 forms the gap junction channels that open a radial diffusion path through non-compact myelin at Schmidt-Lanterman incisures and paranodal loops, and GJB1 whole-gene deletions produce the same typical phenotype as missense alleles. Affected males are hemizygous, with no second allele to lose, and Gjb1-null mice develop the same demyelinating pathology, so restoring connexin 32 in Schwann cells is predicted to rescue. Severity in heterozygous females tracks the proportion of Schwann cells expressing the mutant X.

Rationale: Deletion of the entire gene is as damaging as any missense allele, the cleanest evidence that the lesion is absent channel function. Several mutants are retained in the ER or Golgi and interfere with connexin hemichannels in expression systems, and a dominant-negative action has been argued from that work, but interference is not needed to explain disease in a male with nothing to interfere with. Skewed X-inactivation sets female severity.

Learn more about CMTX1 →

Prediction: The literature predicts a dominant-negative mechanism for CMTDIF: the G-protein beta-4 subunit (Gβ4) functions only in obligate partnership, dimerizing with Gγ and docking onto Gα, and the dominant substitutions K89E and G77R (Soong et al. 2013) yield a subunit that still enters those heterotrimers while failing to transmit receptor signaling in the Schwann cells where GNB4 is strongly expressed. Signaling therefore fails in cells that also carry a normal GNB4 allele.

Rationale: A shortage of Gβ4 is not what these alleles create. The mutant subunit takes its place in the signaling complex and then does not carry the signal through, which is why halving dosage is the wrong model for the disease. The mechanism rests on two substitutions characterized in a single 2013 study, so confidence sits at medium.

Learn more about CMTDIF →

Mechanistic basis: Hypomorphic

Prediction: Enzyme biochemistry predicts a biallelic loss-of-function mechanism for CMT-HADHB: the gene encodes the beta subunit of the mitochondrial trifunctional protein, which carries long-chain 3-ketoacyl-CoA thiolase activity within an alpha/beta hetero-octamer assembled with HADHA, and recessive variants reduce long-chain fatty-acid beta-oxidation. Later-onset alleles retaining partial activity produce the chronic CMT presentation rather than the infantile cardiomyopathic one, a gradient of residual enzyme rather than a gradient of toxicity.

Rationale: Losing the beta subunit destabilizes the assembled complex, so alpha subunit levels fall with it and flux through the final steps of the beta-oxidation spiral drops together. That coordinated depletion, with accumulating long-chain acylcarnitines as its biochemical signature, is the profile of an enzyme deficiency carried on both alleles.

Learn more about CMT-HADHB →

Prediction: Enzymology and the recessive HARS1 phenotypes predict a dominant-negative mechanism for CMT2W: histidyl-tRNA synthetase works as an obligate homodimer, and the heterozygous missense substitutions reduce histidine charging while still permitting dimerization, so a mutant subunit compromises dimers that also contain normal protein. Biallelic HARS1 variants instead cause Usher syndrome type IIIB and a recessive multisystem disease, phenotypes that show what genuine depletion of this enzyme produces.

Rationale: Because subunits pair at random, three quarters of the dimers in a heterozygote contain mutant protein, a shortfall well beyond the half-normal activity that dosage loss alone would predict. The aaRS field has not settled between interference and a toxic gain of function, so the mechanism is graded medium.

Learn more about CMT2W →

Mechanistic basis: Complete loss

Prediction: The literature predicts a biallelic loss-of-function mechanism for CMT-HINT1: missense and truncating variants, including the Slavic founder allele p.Arg37Pro, destabilize the histidine-triad hydrolase and the homodimer it forms, and disease appears only when both copies are compromised. The distinguishing feature is neuromyotonia, with delayed muscle relaxation and myotonic discharges on electromyography accompanying the motor-predominant CMT, and it segregates with the recessive genotype while single-allele carriers stay well.

Rationale: Mutant HINT1 proteins fail to fold and dimerize, so the enzyme is missing from the cell rather than present in an altered form, and no interfering species is left behind. Carrier parents in the large European cohorts show neither neuromyotonia nor CMT, which is the population-scale version of the same result.

Learn more about CMT-HINT1 →

Mechanistic basis: Hypomorphic

Prediction: Russe type CMT4G traces to a Roma founder variant in an alternative untranslated first exon of HK1 rather than in coding sequence, and the literature predicts a purely quantitative loss: the change lowers the HK1 transcript used in peripheral nerve while leaving the ubiquitously expressed hexokinase 1 isoform intact, which is why homozygotes develop CMT without the hemolytic anemia of coding HK1 deficiency. Nerve HK1 expression returned to normal is predicted to correct the deficit.

Rationale: A variant in noncoding sequence can act only through how much protein gets made, and here the nerve-specific transcript is reduced while other tissues keep their supply. Two copies of the founder haplotype are required, and the sparing of erythrocyte hexokinase in these homozygotes marks the defect as quantitative and tissue-restricted.

Learn more about CMT4G →

Mechanistic basis: Neomorphic

Prediction: The literature predicts a neomorphic gain-of-function mechanism for CMT2F: dominant HSPB1 missense substitutions give HSP27 abnormally tight binding to tubulin, microtubules, and neurofilaments and a tendency to form insoluble species, so cargo transport along the long motor and sensory axons affected in CMT2F fails from a new activity rather than from lost chaperone capacity. Mice expressing mutant HSP27 reproduce the phenotype with reduced alpha-tubulin acetylation, while the Hspb1 null mouse is near normal, which places the defect in the mutant protein itself.

Rationale: Hyperactive engagement of the cytoskeleton is the operative lesion, with mutant HSP27 clamping onto microtubules and neurofilaments and stalling the transport distal motor axons depend on, an activity the normal chaperone does not have. These same substitutions also reduce the chaperone capacity of the mixed oligomer, leaving a loss component that cannot be ruled out and holding the grade at medium.

Learn more about CMT2F →

Mechanistic basis: Neomorphic

Prediction: Functional evidence supports a toxic gain-of-function mechanism for dHMN-2B: heterozygous HSPB1 mutants acquire interactions the wild-type small heat-shock protein does not make, binding tubulin and HDAC6, lowering tubulin acetylation, destabilizing neurofilament assembly, and stalling cargo movement along the long motor axons that fail first in dHMN. HDAC6 inhibition restores axonal transport and motor performance in mutant mice, pointing to an acquired activity rather than reduced chaperone dosage. Whether mutant subunits also drag wild-type protein down inside the oligomer is unsettled, and the grade is medium on that point.

Rationale: Tubulin and HDAC6 binding by mutant HSP27 is an activity the normal protein never performs, and the resulting tubulin hypo-acetylation accounts for the transport failure in distal motor axons where these families show weakness with minimal sensory signs. Pharmacological rescue by HDAC6 inhibition is the strongest argument for an acquired toxic function; the competing dominant-negative pull on the chaperone oligomer keeps the mechanism partly open.

Learn more about dHMN-2B →

Mechanistic basis: Unresolved

Prediction: HSPB3 was tied to dHMN-2C by one dominant missense allele, R7S, in essentially a single family (Kolb et al. 2010). Small heat shock proteins act as oligomers and HSPB3 partners with HSPB2, so a mutant subunit could act on its partners rather than simply subtract activity, but the literature offers no HSPB3-specific functional study to distinguish reduced function from a toxic or interfering one.

Rationale: Oligomeric chaperone biology makes an interfering effect the natural hypothesis, and the dominant alleles of the related HSPB1 and HSPB8 encourage it, yet no experiment has examined the R7S protein itself. A hypothesis borrowed from relatives of a gene is not evidence about that gene, which holds confidence low.

Learn more about dHMN-2C →

Mechanistic basis: Neomorphic

Prediction: The evidence supports a toxic gain-of-function mechanism for CMT2L: heterozygous alpha-crystallin-domain substitutions in HSPB8, principally K141N and K141E, make the small heat-shock protein aggregation-prone, and the resulting insoluble species impair protein quality control and autophagic clearance in the long motor and sensory axons that define the CMT2L presentation. A knock-in mouse carrying the human substitution develops the disease, locating the defect in the abnormal protein rather than in the amount of HSPB8 available.

Rationale: Aggregation is an activity wild-type HSPB8 never shows, and the mutant chaperone pulls HSPB1 and other partners into its inclusions, so supplying normal HSPB8 is not predicted to clear them. That same sequestration leaves room for interference with the partner chaperones alongside the toxic gain, which is why confidence stops at medium.

Learn more about CMT2L →

Mechanistic basis: Neomorphic

Prediction: The literature predicts a toxic gain-of-function mechanism for dHMN-2A: the recurrent heterozygous substitutions K141N and K141E sit in the alpha-crystallin domain of HSPB8 and confer an aggregation-prone character that overwhelms chaperone-assisted selective autophagy in motor neurons, with weakness in these families confined to distal motor units and sensory nerves largely spared. Recurrence at a single residue argues against a dosage effect, so restoring wild-type HSPB8 is not predicted to clear the aggregates. Sequestration of partner chaperones leaves a second mechanism in play and confidence at medium.

Rationale: Knock-in mice carrying K141N develop the distal motor phenotype while Hspb8 nulls stay healthy, which separates an acquired toxic property from simple loss of the chaperone. The aggregates also pull HSPB1 and BAG3 out of circulation, a real dominant-negative contribution running alongside the neomorphic one, and that unresolved overlap is why the grade stops short of high.

Learn more about dHMN-2A →

Mechanistic basis: Hypomorphic

Prediction: IGHMBP2 encodes a helicase/ATPase active in RNA metabolism and translation, and CMT2S arises when two recessive alleles reduce that activity without abolishing it. The literature predicts loss of function graded by residual enzyme: allele combinations leaving partial helicase and ATPase output produce this CMT phenotype, whereas near-complete loss of the same protein produces the far more severe allelic disease SMARD1 (dHMN-6), with infantile diaphragmatic paralysis.

Rationale: Genotype-phenotype correlation within IGHMBP2 does the arguing: residual helicase and ATPase activity rises across the SMARD1 to CMT2S range, so the quantity of working enzyme, not the presence of an abnormal one, sets severity. A continuous dose-response of that kind belongs to recessive loss and has no natural reading as interference.

Learn more about CMT2S →

Mechanistic basis: Hypomorphic

Prediction: The literature predicts a biallelic loss-of-function mechanism for dHMN-6 (SMARD1): pathogenic IGHMBP2 variants lower the ATPase and helicase activity of the enzyme, and the residual activity a genotype retains tracks severity, running from infantile diaphragmatic paralysis and respiratory failure to a milder later-onset motor phenotype. The nmd mouse, carrying a hypomorphic Ighmbp2 allele, reproduces motor neuron degeneration, and restored wild-type IGHMBP2 is predicted to rescue.

Rationale: Enzymatic output, not protein identity, sets the phenotype: genotypes retaining measurable helicase activity give later-onset weakness while near-null combinations cause infantile respiratory failure. A graded dose response of that kind follows from a quantitative deficit, and single-variant carriers are unaffected.

Learn more about dHMN-6 →

Mechanistic basis: Overactivity

Prediction: The literature predicts a gain-of-function mechanism for CMTDIE: the dominant missense variants cluster in the N-terminal diaphanous-inhibitory domain and break the intramolecular DID-DAD contact that holds INF2 in its autoinhibited state, leaving the formin constitutively competent to polymerize actin and to drive ER-actin-dependent mitochondrial fission. The same DID allele class also produces focal segmental glomerulosclerosis, which fits an activating lesion rather than reduced INF2 dosage.

Rationale: Releasing the autoinhibitory clamp leaves the mutant formin doing more of INF2's own job, unregulated actin assembly and the mitochondrial fission that depends on it, with no requirement that the wild-type allele be affected. Whether mutant subunits additionally alter normal INF2 within dimers has not been settled, and confidence sits at medium for that reason.

Learn more about CMTDIE →

Mechanistic basis: Gene unknown

Prediction: Rather than a coding mutation, CMTX3 is attributed to a structural rearrangement: a large direct insertion of chromosome 8q24.3 sequence into Xq27.1, der(X)dir ins(X;8)(q27.1;q24.3), segregating with the disease (Brewer et al.). The inserted material lands in a gene desert, so the literature predicts a positional or regulatory effect on a neighboring gene rather than damage to any coding sequence, and no target transcript has been established.

Rationale: Nothing in this genetics points at a protein: the lesion is the rearrangement itself, and a regulatory insult delivered across a gene desert leaves nothing to characterize biochemically while the affected transcript is unnamed. The regulatory target is hypothesized rather than demonstrated, which keeps confidence low.

Learn more about CMTX3 →

Mechanistic basis: Unresolved

Prediction: Published data leave the mechanism of CMT-INSC open: the one reported dominant allele, p.Met70Arg, alters an adaptor that bridges PAR3 to LGN, and it is hypofunctional toward LGN while binding PAR3 more tightly. Weakened LGN engagement reads as loss and tighter PAR3 binding reads as sequestration of a shared partner, while the cross-species rescue in Drosophila does not show which of the two dominates in a human heterozygote.

Rationale: A single substitution that loses one interaction and gains another is the difficulty here, because it fits haploinsufficiency and a sequestration effect equally well, and the two predict opposite outcomes from added wild-type protein. Mutant and wild-type inscuteable have not been co-expressed to see which prevails.

Learn more about CMT-INSC →

Mechanistic basis: Unresolved

Prediction: The reported evidence on CMT1J does not settle a mechanism: recurrent heterozygous ITPR3 missense variants affect a subunit of the homotetrameric IP3 receptor, so one mutant subunit is present in most assembled channels and can pull function well below half. That architecture accommodates suppressed calcium release and a leaky, overactive channel equally well, and no electrophysiology or calcium imaging on the patient alleles has been published.

Rationale: Tetramer assembly is what makes this ambiguous: subunit-level interference and constitutive channel opening are both dominant at the same genotype, yet they move cytosolic calcium in opposite directions. Measuring IP3-evoked release from the mutant alone and alongside wild-type would separate them, and that experiment is missing.

Learn more about CMT1J →

Mechanistic basis: Unresolved

Prediction: The literature leaves CMT2HH mechanistically undefined: the dominant JAG1 missense variants reduce but do not abolish surface Jagged-1 and shift dosage-sensitive Notch signaling in a pattern the primary study set apart from the JAG1 haploinsufficiency behind Alagille syndrome. Jagged-1 activates Notch on neighboring cells while inhibiting Notch within its own, so a partly functional ligand can change the balance between those roles without amounting to a clean loss or gain.

Rationale: That whole-gene deletions and truncating JAG1 alleles produce Alagille syndrome and not CMT argues these missense products do something other than lower ligand supply. Which something, a trans-activation deficit or a shifted cis-inhibitory load, has never been measured in nerve.

Learn more about CMT2HH →

Mechanistic basis: Complete loss

Prediction: Mouse and cell data predict recessive loss in CMTRIE: KCTD11 serves as a substrate adaptor for the Cullin3 E3 ubiquitin ligase, and the recessive truncating and missense variants yield a protein cleared by autophagic degradation, so its substrates go un-ubiquitinated. A Kctd11-knockout mouse reproduces the myelin defect seen in patients, tying the phenotype to absence of the adaptor rather than to any activity of the mutant form.

Rationale: Rapid degradation of the mutant adaptor is the decisive observation: the variant protein does not persist long enough to occupy Cullin3 or to act on the product of the other allele, leaving depletion as the operative lesion. The knockout mouse phenotype carries that argument further. Replication rests on a small number of families, holding the grade short of high.

Learn more about CMTRIE →

Mechanistic basis: Complete loss

Prediction: The literature predicts a biallelic loss-of-function mechanism for HSN-2C: the disease alleles are frameshifts confined to the alternatively spliced KIF1A exon 25b, homozygous in three unrelated families and compound heterozygous in a fourth, so the isoform that hauls dense-core and synaptic vesicle precursors down sensory axons is truncated while canonical kinesin-3 is spared. Heterozygous de novo motor-domain missense variants instead cause a dominant KIF1A-associated neurological disease.

Rationale: Sparing the canonical isoform is what makes two truncating alleles survivable at all, since Kif1a-null mice do not live. The lesion is confined to a splice form the sensory neuron depends on, and carrier parents are neurologically intact, which places the phenotype at a threshold within that isoform rather than at any interference between mutant and normal motor.

Learn more about HSN-2C →

Prediction: The literature predicts a dominant-negative mechanism for CMT-KIF5A: kinesin-1 heavy chain carries cargo as a homodimer with two coordinated motor heads, so the dominant N-terminal motor-domain missense variants place one defective head in a dimer built from both alleles, and that motor still binds microtubules while failing to move cargo instead of simply reducing the number of working motors. Wild-type supplementation is not predicted to rescue while mutant subunits keep pairing with it. The large-deletion dosage evidence cited for this gene is thin, and the mechanism stays debated at low confidence.

Rationale: Kinesin-1 heavy chain moves cargo as a two-headed homodimer, so a single motor-domain mutant is enough to stall a dimer assembled from both alleles, a pattern that fits dominant transmission better than a halved supply of intact motors.

Learn more about CMT-KIF5A →

Mechanistic basis: Neomorphic

Prediction: The literature predicts a toxic gain-of-function mechanism for CMT1C: dominant missense changes in the conserved C-terminal domain of LITAF/SIMPLE cause the protein to misfold, leave the late endosomal membrane, and collect in cytoplasmic aggregates that impede endosome-to-lysosome trafficking and receptor degradation. Litaf-null mice do not develop CMT, which removes haploinsufficiency from consideration and leaves the mislocalized aggregate as the pathogenic entity.

Rationale: The C-terminal domain anchors SIMPLE to the endosomal membrane, and the disease alleles break that anchoring, not the gene's output, so mutant SIMPLE acts from a compartment it does not belong in and added wild-type is not predicted to clear the aggregates already formed. Whether sequestered wild-type protein contributes is unsettled, and medium confidence reflects it.

Learn more about CMT1C →

Mechanistic basis: Hypomorphic

Prediction: Lamin A/C forms the meshwork lining the inner nuclear membrane, and the founder substitution p.R298C, carried homozygously in families of North African origin, lowers functional lamin below what Schwann cells and neurons require. The evidence supports biallelic loss of function in CMT2B1: carriers of a single copy of the identical substitution remain neurologically normal, which places the requirement squarely on losing both copies.

Rationale: Homozygosity for p.R298C is the genetic signature here: one intact LMNA copy sustains normal nerve, two mutant copies do not. The substitution acts as a hypomorph that lowers usable lamin A/C rather than assembling into the lamina and interfering with it, which is how the dominant laminopathies at this locus operate. Its missense nature leaves the residual-activity question open at medium confidence.

Learn more about CMT2B1 →

Mechanistic basis: Repeat expansion

Prediction: For CMT-LRP12, the literature predicts a repeat-expansion gain of function: a heterozygous CGG expansion in the 5' noncoding region of LRP12 is translated by repeat-associated non-AUG initiation into an aggregation-prone polyglycine product, while the coding sequence remains intact and normal LRP12 protein continues to be made. The same expanded locus produces oculopharyngodistal myopathy, an allelic series in which the repeat, not LRP12 function, is what varies.

Rationale: Because the expansion sits outside the coding sequence, nothing is missing from the LRP12 protein, so supplying more wild-type is not predicted to rescue; the pathology comes from what the repeat itself generates. Whether the toxic agent is the expanded RNA or the polyglycine it encodes is still argued, and that caps confidence at medium.

Learn more about CMT-LRP12 →

Mechanistic basis: Mixed

Prediction: CMT2P is defined across two allele classes at LRSAM1 and the literature predicts loss of function for both. Guernsey et al. (2010) established the recessive form, in which biallelic variants strip the RING-domain E3 ubiquitin ligase of activity, and Weterman et al. (2012) established a dominant form built on a frameshift segregating through a large family. Neither route produces an enzyme that does something new; both subtract ligase activity from the neuron.

Rationale: One gene, two inheritance patterns, one direction of effect. Because the dominant frameshifts stop short of the RING domain while retaining the coiled-coil self-association region, whether the dominant form works by halved dose alone or by the truncated product engaging the intact one has not been separated, and that open question keeps confidence low.

Learn more about CMT2P →

Mechanistic basis: Hypomorphic

Prediction: Cohort evidence indicates a biallelic loss-of-function mechanism for CMT-MCM3AP: homozygous and compound-heterozygous variants lower the amount or activity of GANP, the TREX-2 component that docks maturing messenger RNA at the nuclear pore for export, and patients present with early-onset sensorimotor CMT, often with intellectual disability. Reported genotypes always retain some functional GANP, since at least one allele is hypomorphic rather than null, matching the essentiality of nuclear mRNA export.

Rationale: Every reported genotype preserves residual GANP, placing the disease on a dosage gradient beneath a threshold that complete absence would fall straight through. What remains unexplained is why peripheral neurons in particular fail when a housekeeping export complex is only partly depleted, and that gap in the mechanism is why the grade stops short of high.

Learn more about CMT-MCM3AP →

Prediction: The literature predicts a dominant-negative mechanism for CMT2A: mitofusin-2 tethers and fuses adjacent mitochondrial outer membranes by oligomerizing in trans with MFN2 and MFN1, so a mutant GTPase entering that tethering complex blocks fusion of both mitochondria rather than withdrawing half the fusion capacity. Mitochondrial fusion, transport, and tethering all fail in the presence of the mutant, and the early-onset severity of several alleles exceeds what a halved MFN2 supply produces. Biallelic MFN2 alleles instead cause a separate recessive subtype.

Rationale: Fusion requires two membranes bridged by one oligomer, which is what lets a single mutant subunit disable an event needing protein contributed from both mitochondria. That MFN1 can complement certain MFN2 mutants in trans, and that reduced MFN2 dosage remains a live explanation for milder alleles, leaves the mechanism genuinely contested at medium confidence.

Learn more about CMT2A →

Mechanistic basis: Hypomorphic

Prediction: The literature predicts a biallelic loss-of-function mechanism for CMT2A2B: mitofusin-2 tethers outer mitochondrial membranes for fusion, and the recessive MFN2 genotypes combine alleles that lower that activity rather than abolish it, with heterozygous relatives unaffected. Complete mitofusin-2 loss is not compatible with this phenotype, since Mfn2-null mice die in midgestation and near-null human genotypes present as lethal neonatal disease, so at least one partial-function allele is required.

Rationale: These families sit at a threshold rather than at zero: enough fusion activity survives to carry development, and too little survives to maintain the longest axons into adult life. Homozygous p.Arg707Trp is the recurring genotype, a reduced-function allele rather than a null. Nothing here is positioned to act on a wild-type partner, which is what separates this from the dominant MFN2 subtype.

Learn more about CMT2A2B →

Mechanistic basis: Hypomorphic

Prediction: For CMT2B4 the literature predicts a biallelic loss of mitofusin-2 activity, carried by homozygous or compound-heterozygous MFN2 alleles that reduce outer-membrane tethering and fusion below what peripheral nerve tolerates while leaving carriers with a working copy's worth. Mitofusin-1 compensates for diminished mitofusin-2 in tissues that express it abundantly, an asymmetry invoked to explain why peripheral nerve is selectively vulnerable.

Rationale: Two null MFN2 alleles are incompatible with survival, so the genotypes behind this subtype necessarily retain partial fusion activity, and the disease marks where that remainder falls short in the longest axons. That is the point of separation from the dominant subtype, where one missense allele suffices because the abnormal protein is built into the fusion machinery rather than diminished within it.

Learn more about CMT2B4 →

Prediction: The literature predicts a dominant-negative mechanism for HMSN-6A: mitofusin-2 fuses apposed mitochondrial outer membranes only as an oligomer acting in trans, so heterozygous MFN2 missense alleles seed complexes that also contain normal MFN2 and MFN1, blunting fusion and the axonal transport of mitochondria. Wild-type add-back is not predicted to restore fusion in complexes that already hold a mutant subunit. Biallelic MFN2 alleles instead cause a severe early-onset recessive loss-of-function disease. That MFN1 co-expression rescues several mutants in trans keeps the mechanism contested and confidence at medium.

Rationale: Mutant mitofusin-2 enters the same trans tethering complexes as normal MFN2 and MFN1, and because tethering requires competent oligomers on both membranes, one defective subunit stalls the reaction rather than merely halving capacity. Mitochondria then clump and move poorly along the axon, a deficit the optic nerve and the longest peripheral axons register first.

Learn more about HMSN-6A →

Mechanistic basis: Mixed

Prediction: The literature predicts loss of function for CMT2T across both of its allele classes. Higuchi et al. (2016) established the recessive form, where homozygous and compound-heterozygous MME variants abolish neprilysin, the membrane metalloendopeptidase the gene encodes, and Auer-Grumbach et al. (2016) established a late-onset dominant form carried by rare heterozygous alleles. Returning neprilysin activity is predicted to address either.

Rationale: Disease follows from neprilysin substrates going uncleaved rather than from an altered enzyme acquiring a target. The recessive form is the cleaner demonstration, since two null alleles leave no activity at all; how a single defective copy reaches threshold in the late-onset dominant form is characterized clinically more than biochemically, and confidence sits at medium on that account.

Learn more about CMT2T →

Mechanistic basis: Overactivity

Prediction: The literature predicts an overactivity gain-of-function mechanism for CMT2Z: dominant MORC2 missense variants cluster in and around the GHKL ATPase module, where they release the autoinhibited state of the protein and leave the ATPase constitutively engaged. The consequence is excessive HUSH-complex-directed H3K9me3 deposition and over-silencing of target loci. Structural and cellular work converge on hyperactivation rather than reduced MORC2 function, and the recurrent p.Arg252Trp allele produces the severe early-onset end of the CMT2Z spectrum.

Rationale: Relief of autoinhibition explains the phenotype directly: the mutant ATPase runs without its normal restraint and the HUSH complex silences more than it should, an excess of native activity rather than a shortfall of it. Because the damage comes from what the mutant protein does, not from what it prevents the wild-type from doing, neither haploinsufficiency nor interference accounts for the dominant transmission.

Learn more about CMT2Z →

Mechanistic basis: Hypomorphic

Prediction: MPV17 is an inner mitochondrial membrane protein that sustains mitochondrial DNA copy number through the supply of deoxyribonucleotides. Evidence in CMT2EE predicts loss of that function across both MPV17 copies, with mtDNA depletion in post-mitotic tissue as the downstream lesion. Severe alleles of the same gene produce hepatocerebral mtDNA depletion syndrome and Navajo neurohepatopathy, placing this CMT at the mild end of one allelic series.

Rationale: Alleles retaining partial MPV17 function spare liver and brain and surface instead as peripheral disease in adulthood, while null combinations present in infancy with hepatocerebral failure. A graded relation of that kind between residual protein and severity is the fingerprint of a recessive dosage deficit rather than a toxic product.

Learn more about CMT2EE →

Mechanistic basis: Mixed

Prediction: In CMT1B the literature supports two mechanisms carried by different MPZ alleles. Most dominant missense alleles, S63del and R98C among them, produce P0 that misfolds and is retained in the endoplasmic reticulum, driving chronic unfolded protein response activation and disrupting the homophilic P0 adhesion lattice of compact myelin, so added wild-type P0 is not predicted to fully rescue. A separate set of MPZ-deficiency alleles causes a milder loss-of-function CMT, matching the late-onset demyelination of heterozygous-null mice.

Rationale: Retention of misfolded P0 in the endoplasmic reticulum is a toxic event rather than a shortfall of protein, and the adhesion lattice of compact myelin tolerates mutant subunits poorly, which is what makes the missense alleles dominant. The haploinsufficiency arm rests on rarer MPZ-deficiency alleles and the heterozygous-null mouse, a thinner evidence base than the misfolding arm, and medium confidence reflects that asymmetry.

Learn more about CMT1B →

Prediction: Evidence at the MPZ locus supports a dominant-negative mechanism for CMT2I: the late-onset heterozygous missense alleles, among them Thr124Met, yield a P0 glycoprotein that is still trafficked into compact myelin, where it occupies the adhesive lattice and degrades adhesion between the wrapped membranes rather than being absent from them. A residual toxic gain of function has not been ruled out for this allele class, and the medium grade reflects that.

Rationale: Allele identity, not P0 dosage, sets the presentation at this locus: substitutions in the extracellular domain give outcomes ranging from severe infantile demyelination to this late-onset axonal disease, a spread that reduced protein levels cannot generate. Mutant P0 has to reach myelin to produce that, and once embedded it interferes with the wild-type molecules packed around it.

Learn more about CMT2I →

Prediction: The literature predicts a dominant-negative mechanism for CMT2J: P0 oligomerizes to hold adjacent wraps of compact myelin together, and the classic Thr124Met allele is trafficked into that lattice, so mutant and normal subunits mix within a single adhesive array. Carriers show adult-onset axonal degeneration, often with pupillary abnormality and hearing loss, at a severity that MPZ null and deficiency alleles, which merely halve P0, do not reach.

Rationale: Homophilic adhesion is a cooperative property of P0 oligomers, so an array built from a mixture of Thr124Met and normal subunits fails at a level neither sets alone. That cooperativity, not the quantity of protein made, is what the phenotype tracks, and whether the mutant additionally acquires a toxic activity in Schwann cells remains argued.

Learn more about CMT2J →

Prediction: Clinical and structural evidence supports a dominant-negative mechanism for CMTDID: heterozygous MPZ substitutions that misfold the extracellular domain still deliver P0 to the Schwann cell membrane, where the abnormal molecule keeps the normal ones around it from packing compact myelin correctly. Whether misfolded P0 additionally injures Schwann cells through unfolded-protein-response stress, a toxic route rather than an interfering one, is unsettled, which keeps the grade at medium.

Rationale: Intermediate conduction slowing is itself informative: it describes myelin that gets built and then performs badly, which is what a defective adhesion molecule sitting in the membrane produces and not what a simple shortage of P0 produces. Severity across this allele class tracks how badly a given substitution perturbs the lattice.

Learn more about CMTDID →

Mechanistic basis: Hypomorphic

Prediction: The literature attributes CMT-ATP6 to loss of function: maternally transmitted MT-ATP6 missense variants such as m.9185T>C impair proton translocation through subunit a of the ATP synthase Fo domain, lowering ATP output in peripheral nerve. Pathogenic alleles in this presentation are typically homoplasmic or carried at high heteroplasmy, and severity tracks the proportion of mutant mitochondrial genomes, the threshold behaviour familiar from the m.8993T>G NARP allele at the same gene.

Rationale: Substituted residues in subunit a line the proton half-channel that couples rotation to ATP synthesis, so the deficit is energetic rather than toxic. Because mutant subunit a is still built into an assembled ATP synthase, how much of the impairment reflects lost proton conduction and how much reflects poor coupling within the complex is not settled, and the confidence grade reflects that unresolved point.

Learn more about CMT-ATP6 →

Mechanistic basis: Complete loss

Prediction: MTMR2 encodes a phosphoinositide 3-phosphatase that turns over PI3P and PI(3,5)P2 at Schwann cell membranes, and the literature predicts that recessive CMT4B1 alleles remove that catalytic activity outright: nonsense and frameshift variants delete the protein, while catalytic-domain missense variants inactivate the phosphatase site. Mtmr2-null mice develop the same focally folded myelin found on patient biopsy, and reinstating phosphatase activity is predicted to rescue.

Rationale: Without MTMR2 the membrane phosphoinositide pool is never dephosphorylated, and the sheath folds back on itself: the outfoldings on biopsy read out unopposed PI(3,5)P2 signaling. Null and catalytic-dead alleles converge on the same absent enzyme activity, and parents heterozygous for either have normal nerves.

Learn more about CMT4B1 →

Mechanistic basis: Complete loss

Prediction: Published evidence points to a biallelic loss-of-function mechanism for CMT-MTRFR: MTRFR (C12orf65) encodes a GGQ-domain member of the mitochondrial release-factor family that acts in rescue and recycling of stalled mitoribosomes rather than in standard stop-codon termination. The reported truncating variants remove that activity outright, mitochondrial protein synthesis stalls, and combined respiratory-chain deficiency follows in long motor and optic axons. Because the protein is simply absent, restored wild-type is predicted to rescue.

Rationale: Loss of the recycling factor leaves stalled mitoribosomes sequestered and depletes the pool available for translation, starving the respiratory chain of mtDNA-encoded subunits. Truncating alleles on both copies, with unaffected heterozygous parents, place the defect in absent protein rather than in a mutant product. Reported cohorts stay small and phenotypically mixed across optic atrophy, spastic paraparesis, and CMT, which keeps confidence at medium.

Learn more about CMT-MTRFR →

Mechanistic basis: Unresolved

Prediction: MYH14 encodes non-muscle myosin heavy chain IIC, and heterozygous missense variants in it are the established cause of dHMN-MYH14, though the literature has not pinned down the lesion. The protein dimerizes through a coiled-coil rod and assembles into bipolar filaments, an architecture in which a mutant heavy chain could compromise filaments built with normal subunits, but no allele-specific functional work separates that possibility from simple loss.

Rationale: Filament assembly is what makes interference plausible, since co-assembling subunits give a mutant chain access to structures containing the wild-type protein. Plausibility is not data: with no cell or biochemical work on these variants, and MYH14 also implicated in dominant hearing loss, the mechanism stays open at low confidence.

Learn more about dHMN-MYH14 →

Mechanistic basis: Unresolved

Prediction: Current evidence assigns no mechanism to CMT-MYO9B: an ultra-rare 2023 association links heterozygous MYO9B variants to dominant CMT, and myosin IXb is an unusual protein to reason about, a processive actin-based motor that carries a RhoGAP domain in its tail and switches RhoA off as it travels. Such a variant could lower total GAP output or deliver an intact GAP domain to the wrong place along the cytoskeleton, and no functional work has been done.

Rationale: Because motor activity and Rho regulation sit in one polypeptide, the amount of RhoGAP output and its placement can fail independently, and each failure implies a different mechanism. Neither RhoA activity nor myosin IXb localization has been measured in cells from these patients, so neither route can be excluded.

Learn more about CMT-MYO9B →

Mechanistic basis: Unresolved

Prediction: The literature does not resolve a mechanism for CMT2V: heterozygous NAGLU missense variants cause a late-onset dominant CMT (Tetreault et al. 2015), yet the obligate carriers of the biallelic NAGLU-null alleles that cause Sanfilippo B are neurologically unaffected, which argues against simple haploinsufficiency. No toxic gain-of-function or dominant-negative effect of the missense alpha-N-acetylglucosaminidase has been demonstrated, so the dominant mechanism is undetermined.

Rationale: One working allele of NAGLU is sufficient for normal peripheral nerve function, as the unaffected parents of children with Sanfilippo B show, so reduced dosage alone cannot explain a dominant CMT. What the missense enzyme does instead, whether it interferes with the residual wild-type protein or acquires a toxic property, has never been tested, and confidence stays low.

Learn more about CMT2V →

Mechanistic basis: Hypomorphic

Prediction: Available evidence supports biallelic loss of function in CMT-NAMPT: NAMPT catalyzes the rate-limiting step of the NAD+ salvage pathway, converting nicotinamide to nicotinamide mononucleotide, and the reported homozygous variants lower that enzymatic output. Axons depend on continuous local NAD+ resynthesis to hold off the programmed degeneration that follows NAD+ collapse, so reduced salvage flux leaves the longest fibers metabolically exposed. Only a handful of families have been described, in 2025, and confidence stays low on that basis.

Rationale: NAD+ supply, not a mutant enzyme species, is what fails: the deficit is quantitative, and recessive segregation with unaffected heterozygous parents shows that roughly half-normal activity suffices. Supplying functional enzyme, or precursors entering below the block, is the corrective direction the biochemistry points to. The evidence rests on very few reported families, which is where the low grade comes from.

Learn more about CMT-NAMPT →

Prediction: The evidence supports a dominant-negative mechanism for CMT-NARS1: asparaginyl-tRNA synthetase works as a homodimer, and the heterozygous CMT variants are catalytically null or hypomorphic yet still assemble with their normal partner, which 2025 functional studies show impairs that partner rather than only halving aminoacylation. Biallelic true loss of NARS1 causes a neurodevelopmental disorder, not CMT.

Rationale: A catalytically dead subunit that still dimerizes costs the cell more than one missing copy of the enzyme, and the two 2025 studies show that interference on the wild-type NARS1 protein directly. Whether the dominant synthetase alleles instead acquire a toxic activity is still argued, which is why the grade sits at medium.

Learn more about CMT-NARS1 →

Mechanistic basis: Complete loss

Prediction: In the Roma founder disease HMSN-Lom, the homozygous NDRG1 p.Arg148* allele truncates the transcript and leaves no detectable NDRG1 protein in patient nerve, and the literature predicts that this absence on both alleles is the whole of the CMT4D mechanism. NDRG1 supports Schwann cell myelin maintenance, and its loss gives the early demyelinating course together with the sensorineural hearing loss that emerges in adulthood in Lom.

Rationale: A protein absent from patient tissue has nothing left to act on, and what remains is a Schwann cell unable to maintain a sheath it originally built normally, which is why early motor milestones are unremarkable before demyelination sets in. Heterozygous carriers of the founder allele across the Roma population are healthy.

Learn more about CMT4D →

Mechanistic basis: Hypomorphic

Prediction: The literature reads CMT-NDUFS6 as a biallelic loss-of-function disease: NDUFS6 encodes a small nuclear-encoded subunit of the N module of complex I, where NADH oxidation feeds the respiratory chain. Recessive variants reduce or abolish its incorporation, leaving complex I under-assembled and its catalytic activity low in tissues with high oxidative demand. Supplying intact subunit is predicted to restore assembly. Reports linking this gene to CMT date only from 2024 and remain few, so confidence sits at low.

Rationale: A structural subunit that is missing or unstable fails to join the assembling complex; it does not occupy a position in the holoenzyme from which it could compromise the wild-type product, which is the specific reason interference is not the better reading here. Complete NDUFS6 loss produces severe infantile complex I deficiency, placing the CMT alleles at the hypomorphic end of one gradient.

Learn more about CMT-NDUFS6 →

Mechanistic basis: Neomorphic

Prediction: The literature predicts a neomorphic gain-of-function mechanism for CMT2CC: heterozygous frameshifts in the NEFH tail domain carry translation past the normal stop codon into the 3'UTR, appending a cryptic amyloidogenic sequence that is never otherwise translated. The resulting neurofilament heavy chain aggregates and drives axonal degeneration, a protein the genome does not normally make, not a shortage of the one it does.

Rationale: The pathogenic element is encoded in the 3'UTR, sequence read only because the frameshift removes the stop codon, so what the allele produces is genuinely novel, not truncated or reduced. The supporting literature is narrow and mutant filaments may also entrap wild-type NF-H, two reasons the mechanism is graded medium and not high.

Learn more about CMT2CC →

Prediction: The literature predicts a dominant-negative mechanism for CMT1F: neurofilament light chain is the obligate backbone subunit of the neurofilament, and mutant NF-L from heterozygous NEFL missense variants incorporates into growing filaments alongside wild-type subunits, which then fail to form a normal network, aggregating in the cell body and stalling neurofilament transport. One mutant subunit therefore compromises filaments made from both alleles, which is why a single variant suffices and why CMT1F slows conduction velocities into the demyelinating range.

Rationale: Filament assembly is cooperative, so the mutant subunit is not absent from the polymer but built into it, blocking network formation and producing perikaryal aggregates in cell models. Halved NF-L supply does not reproduce that pattern, which is what separates this mechanism from loss of function.

Learn more about CMT1F →

Mechanistic basis: Complete loss

Prediction: Neurofilament light chain is the obligate subunit that nucleates neurofilament assembly and sets axonal caliber. Nonsense and frameshift NEFL alleles on both chromosomes leave essentially no NFL protein, and the literature predicts that absence, rather than any abnormal subunit, as the lesion in CMT2B5. The mechanism is subtraction of a structural protein: nothing mutant is deposited in the cytoskeleton for a normal copy to contend with.

Rationale: Near-total absence of neurofilament light chain is a different lesion from the one behind the dominant NEFL subtypes. The missense alleles of CMT1F, CMT2E, and CMTDIG generate a subunit that still enters filaments and disrupts the network from within, which is why one copy is enough to cause disease; losing both copies simply removes the protein.

Learn more about CMT2B5 →

Prediction: The evidence supports a dominant-negative mechanism for CMT2E: dominant NEFL missense alleles yield a neurofilament light chain that co-assembles with wild-type subunits and disrupts the shared filament network, causing perikaryal aggregation and impaired axonal transport. Mice lacking Nefl altogether have a comparatively mild phenotype, while mutant NF-L expressed alongside the normal protein drives motor neuron degeneration, so absence of the subunit and presence of the mutant subunit are not equivalent.

Rationale: Deleting the gene and expressing the mutant give different outcomes in mice, the mutant being far worse, which is the signature of interference rather than reduced supply. Human genetics agrees: biallelic NEFL nulls cause the recessive subtype CMT2B5, a separate disease from this dominant one.

Learn more about CMT2E →

Prediction: Functional and genetic evidence supports a dominant-negative mechanism for CMTDIG: neurofilaments are obligate copolymers nucleated by the light chain, and mutant NF-L from the dominant NEFL alleles co-assembles with normal subunits, blocks filament formation, and forms aggregates that stall axonal transport. Wild-type subunits supplied alongside are drawn into those aggregates rather than rescuing assembly. Biallelic NEFL null alleles instead cause the recessive subtype CMT2B5, a true loss of light chain.

Rationale: Cooperative assembly is why one abnormal subunit costs more than a missing one: mutant NF-L is incorporated at the nucleation step and takes normal subunits with it, so a neuron loses far more filament than a halved gene dose would cost. Patients who make no light chain at all have the distinct recessive disease CMT2B5.

Learn more about CMTDIG →

Mechanistic basis: Hypomorphic

Prediction: At the ligand end of that same NGF/TrkA axis, the literature predicts biallelic loss of function for HSAN-5: the classic homozygous NGF allele p.R100W impairs proNGF processing and secretion of mature NGF, starving small-fiber nociceptors of the neurotrophic support they depend on. Because the defect sits upstream of the receptor and is partial, pain sensation is lost while sweating is largely preserved, which distinguishes HSAN-5 from receptor-level abolition in HSAN-4.

Rationale: A hypomorphic ligand rather than a dead receptor accounts for the narrower phenotype: enough signaling survives for sympathetic function while nociceptor support fails. Restored wild-type NGF is predicted to rescue, and unaffected heterozygous relatives show that one normal allele delivers sufficient ligand.

Learn more about HSAN-5 →

Mechanistic basis: Repeat expansion

Prediction: The literature supports a repeat-expansion gain-of-function mechanism for CMT-NOTCH2NLC: a heterozygous GGC expansion in the 5'UTR drives repeat-associated non-AUG translation of a polyglycine protein, and both the expanded transcript and that product are toxic. The signature here is pathological rather than biochemical, the ubiquitin-positive eosinophilic intranuclear inclusion shared with neuronal intranuclear inclusion disease, seen in patients whose presentation is CMT. Toxicity comes from what the expansion makes, not from reduced NOTCH2NLC activity.

Rationale: One expanded allele fills nuclei with inclusions, and those inclusions contain the polyglycine translation product, so the pathology is assembled from a molecule the repeat creates and the genome does not otherwise encode. Phenotype across the NOTCH2NLC spectrum follows the expanded repeat, not any shortfall of normal protein.

Learn more about CMT-NOTCH2NLC →

Mechanistic basis: Complete loss

Prediction: NGF signaling through the TrkA receptor is the axis at issue, and the literature predicts biallelic loss of function for HSAN-4: recessive NTRK1 alleles spanning nonsense, frameshift, splice, and kinase-domain missense changes abolish TrkA tyrosine kinase activity, so NGF-dependent nociceptive and sympathetic neurons fail to survive development. Absent sympathetic innervation of sweat glands is what adds anhidrosis to the pain insensitivity on this receptor half of the axis.

Rationale: Kinase-dead and truncated TrkA receptors cannot transduce NGF, and heterozygous parents in the reported consanguineous families are asymptomatic, so one intact receptor allele supports normal nociceptor development. Reinstating catalytically competent TrkA is predicted to restore signaling, since the deficit is absent activity rather than a receptor that blocks its wild-type counterpart.

Learn more about HSAN-4 →

Mechanistic basis: Overactivity

Prediction: Published functional work supports an overactivity gain of function in CMTX6: the recurrent PDK3 p.Arg158His substitution tightens the kinase's docking onto the L2 lipoyl domain of the pyruvate dehydrogenase complex, so mutant PDK3 phosphorylates the E1 alpha subunit more efficiently and holds pyruvate dehydrogenase in its inactive state. Chronic suppression of pyruvate entry into the TCA cycle starves long motor and sensory axons of ATP, and hemizygous males are affected more severely than heterozygous females.

Rationale: A single recurrent allele carries this mechanism, and its effect is directional: p.Arg158His raises PDK3 activity toward its normal substrate rather than abolishing it, so the defect is too much inhibitory phosphorylation of pyruvate dehydrogenase, not too little kinase. Resting on one mutation in a small number of families, the mechanism is well argued but narrowly sampled, which is why confidence stops short of high.

Learn more about CMTX6 →

Mechanistic basis: Hypomorphic

Prediction: Cofactor biochemistry predicts that HMSN-6C arises from loss of function at both PDXK alleles: PDXK encodes pyridoxal kinase, which phosphorylates dietary vitamin B6 to pyridoxal 5'-phosphate, the cofactor for scores of enzymes including those of neurotransmitter and sphingolipid metabolism. The recessive variants reduce catalytic activity and lower circulating PLP (Chelban et al. 2019). Giving PLP directly, bypassing the missing kinase step, raises levels and improves some patients, which is the behavior of a deficiency state.

Rationale: Because supplying the product of the blocked reaction corrects the biochemistry, the defect sits in catalysis itself and not in any property the mutant enzyme acquires. Where mitochondrial dynamics drives optic atrophy with peripheral disease elsewhere, this one is a treatable cofactor shortage, and residual kinase activity accounts for the clinical range.

Learn more about HMSN-6C →

Mechanistic basis: Complete loss

Prediction: The literature predicts a biallelic loss-of-function mechanism for HMSN-4 (Refsum disease): PHYH encodes phytanoyl-CoA hydroxylase, the peroxisomal enzyme that opens alpha-oxidation of phytanic acid, and the recessive alleles leave little or no measurable activity in patient fibroblasts. Phytanic acid accumulates in plasma and tissue, and dietary restriction lowers both the stored substrate and the clinical burden. A minority of Refsum disease arises instead from biallelic PEX7 variants that fail to import the enzyme into the peroxisome.

Rationale: Accumulated substrate, not an aberrant enzyme, drives the phenotype, since assays show the hydroxylase activity is simply gone and lowering dietary intake relieves the disease. Two routes converge on the same deficiency, one in the enzyme and one in its import, which locates the lesion in how much activity reaches the peroxisome.

Learn more about HMSN-4 →

Mechanistic basis: Complete loss

Prediction: Published families with CMTRIC point to a biallelic loss-of-function mechanism: PLEKHG5 encodes a guanine nucleotide exchange factor for Rho-family GTPases, and the reported variants, the founder missense p.Arg204Trp together with truncating frameshifts, reduce or abolish that exchange activity. Relatives carrying one variant are unaffected. Other biallelic PLEKHG5 alleles cause a recessive distal spinal muscular atrophy, so the disease alleles at this locus are loss alleles across presentations.

Rationale: Exchange activity toward Rho-family GTPases is reduced or absent, and a frameshift that removes the catalytic machinery produces the same phenotype as the founder missense allele, which is the signature of a shared floor of lost function. Only a handful of families have been characterized, so the genetics carry more weight here than direct functional work, and the grade reflects that gap.

Learn more about CMTRIC →

Mechanistic basis: Complete loss

Prediction: For dSMA-4, the literature predicts biallelic loss of function: PLEKHG5 is a Rho guanine-nucleotide exchange factor that motor nerve terminals use to drive autophagy of synaptic vesicles, and the recessive truncating and destabilizing missense alleles cut both GEF activity and protein stability on each copy. Plekhg5-null mice degenerate at the neuromuscular junction. Other biallelic PLEKHG5 genotypes produce a recessive intermediate form of Charcot-Marie-Tooth disease (CMT), placing dSMA-4 on an allelic spectrum.

Rationale: Two damaged copies are needed because the GEF activity supplied by one intact allele sustains synaptic vesicle autophagy in motor terminals. Whether the missense alleles are complete nulls or isoform-selective hypomorphs has not been settled, and that gap in the allele-level data is what the medium grade reflects.

Learn more about dSMA-4 →

Mechanistic basis: Neomorphic

Prediction: In CMT1G the literature predicts a toxic gain of function: heterozygous PMP2 missense variants such as I43N and T51P destabilize the fatty-acid-binding fold of P2 while leaving it competent to enter compact myelin, where the altered protein perturbs membrane stacking and compaction. Raised PMP2 levels are themselves demyelinating, which fits a mechanism driven by the presence of an abnormal protein in myelin rather than by its absence.

Rationale: Mice lacking P2 build serviceable myelin, so a dominant phenotype cannot rest on a shortage of it. What these alleles contribute instead is a destabilized protein that still reaches the myelin membrane and compromises its packing. Only a few families and limited functional work support the mechanism, so confidence stays low.

Learn more about CMT1G →

Mechanistic basis: Dosage

Prediction: The literature strongly supports a gene-dosage mechanism for CMT1A: a 1.5 Mb tandem duplication at 17p11.2 spanning PMP22 leaves most patients with three functional copies, and overexpression of a structurally normal protein impairs Schwann-cell myelination. Rodent models reproduce the demyelinating phenotype in proportion to PMP22 copy number. The reciprocal 1.5 Mb deletion of the same interval causes HNPP, so one locus yields opposite diseases from too much and too little PMP22.

Rationale: Three copies of PMP22, not a damaged one, is the lesion: the protein made from the duplication is normal in sequence, so the pathology is quantitative and it is the amount that must come down. PMP22 missense mutations produce a third mechanism at the same gene, the dominant-negative misfolding of CMT1E, which is why duplication, deletion and point mutation give diseases that are not variations on each other.

Learn more about CMT1A →

Prediction: The evidence supports a dominant-negative mechanism for CMT1E: heterozygous PMP22 point mutations, the class modeled by the Trembler and Trembler-J alleles, produce a misfolded protein retained in the endoplasmic reticulum, where it aggregates and sequesters wild-type PMP22 together with its interacting partners rather than simply going missing from compact myelin. Because the normal protein is trapped in those aggregates, supplying more of it is not predicted to restore myelination.

Rationale: One deleted PMP22 allele causes HNPP and one extra copy causes CMT1A, yet a single point mutation produces disease more severe than either, which puts CMT1E outside that dosage series. Retention of mutant and wild-type protein in the endoplasmic reticulum, seen in the Trembler models, accounts for the excess severity.

Learn more about CMT1E →

Mechanistic basis: Haploinsufficiency

Prediction: HNPP sits opposite CMT1A on the same 17p11.2 dosage axis, and the literature predicts haploinsufficiency: the recurrent 1.5 Mb deletion removes one PMP22 copy, the exact segment that is duplicated in CMT1A, and private nonsense and frameshift alleles that inactivate one copy produce the same tomaculous phenotype. Half-normal PMP22 is therefore sufficient to cause disease, and restoring PMP22 dosage is predicted to rescue.

Rationale: Deletion and duplication of one 1.5 Mb interval producing opposite myelin diseases is the cleanest dosage experiment in CMT genetics, and it places HNPP on the reduced side: one functional PMP22 copy, no toxic product. Truncating point alleles that phenocopy the deletion confirm that the lesion is quantity of PMP22 rather than a mutant protein acting on the wild-type.

Learn more about HNPP →

Mechanistic basis: Hypomorphic

Prediction: Polynucleotide kinase 3'-phosphatase, the bifunctional enzyme encoded by PNKP, trims 5'-hydroxyl and 3'-phosphate DNA ends so that ligases can seal single- and double-strand breaks. The literature predicts a biallelic loss-of-function mechanism for CMT2B2, in which two hypomorphic or damaging alleles leave kinase and phosphatase output too low to keep pace with break repair in long-lived neurons. Restored wild-type PNKP is predicted to rescue.

Rationale: The PNKP allelic spectrum carries the argument: null and severe alleles produce microcephaly with early-onset seizures or ataxia with oculomotor apraxia, while alleles retaining partial DNA-end-processing activity surface at the milder CMT end. The founder allele p.Gln517* illustrates the point, since it removes only the last few residues and lowers output rather than abolishing it despite reading as a nonsense change.

Learn more about CMT2B2 →

Mechanistic basis: Hypomorphic

Prediction: In CMT-POLG the literature attributes disease to biallelic loss of function: POLG encodes the catalytic subunit of DNA polymerase gamma, the only polymerase that replicates mitochondrial DNA, and recessive alleles such as p.Ala467Thr and p.Trp748Ser leave a hypomorphic enzyme with reduced processivity and impaired coupling to its accessory subunit. Mitochondrial DNA is depleted and accumulates deletions, and the resulting respiratory-chain failure reaches sensory and motor axons alongside the ataxic and epileptic presentations of the POLG spectrum.

Rationale: Two hypomorphic copies are required because partial polymerase activity is tolerated: p.Ala467Thr heterozygotes are common in northern European populations and healthy. Dominant POLG alleles do exist, clustering in the polymerase domain and causing progressive external ophthalmoplegia by interfering with the wild-type enzyme, a different lesion from these recessive hypomorphs. Residual activity varies widely across recessive genotypes, and the grade of medium follows from that.

Learn more about CMT-POLG →

Prediction: The evidence predicts a dominant-negative mechanism for CMT1I: the CMT-associated POLR3B variants cluster in the regions that form the catalytic core of RNA polymerase III, and a subunit carrying such a change assembles into the 17-subunit enzyme and leaves the whole complex impaired. Two hypomorphic POLR3B alleles are required for the recessive leukodystrophy phenotype, so a single allele causing dominant CMT points to something beyond reduced subunit supply.

Rationale: Pol III is assembled once from whatever subunits are on hand, so a defective second-largest subunit takes an entire complex out of service instead of lowering how many complexes exist. That interference is inferred from the enzyme's architecture rather than demonstrated in assay, and the grade is medium for that reason.

Learn more about CMT1I →

Mechanistic basis: Complete loss

Prediction: The literature predicts a biallelic loss-of-function mechanism for HSAN-8: recessive PRDM12 changes, including truncating alleles, PR/SET-domain missense substitutions, and expansions of the C-terminal polyalanine tract, remove the transcriptional and chromatin-modifying activity required to specify and maintain nociceptive sensory neurons. Affected individuals are born without pain sensation because the nociceptor lineage never forms properly, and reinstating PRDM12 activity is predicted to correct the deficit.

Rationale: Polyalanine expansions are the alleles that would most plausibly act by aggregation, yet they deplete PRDM12 from the nucleus rather than trapping the wild-type product with it, and expansion carriers are unaffected. Strict recessive segregation across consanguineous families completes the case for absent nociceptor-specifying activity.

Learn more about HSAN-8 →

Mechanistic basis: Hypomorphic

Prediction: Enzyme measurements across the PRPS1 spectrum support a loss-of-function mechanism for CMTX5: PRS-I generates phosphoribosyl pyrophosphate for purine and pyrimidine synthesis, and the CMTX5 missense alleles are hypomorphic, leaving residual activity that tracks inversely with severity from nonsyndromic deafness through CMTX5 to Arts syndrome. A male's single X carries no reserve, which is why hemizygotes are affected and carrier females largely spared. Alleles that instead raise enzyme activity cause PRPP synthetase superactivity, a gain-of-function disorder with hyperuricemia.

Rationale: Two directions of mutation at one locus separate the mechanisms cleanly: reduced PRS-I activity gives the CMTX5 phenotype, while activating alleles that escape allosteric feedback drive purine overproduction. Graded residual activity, not a distinct behavior of the mutant enzyme, determines where a patient falls along the deficiency spectrum.

Learn more about CMTX5 →

Mechanistic basis: Complete loss

Prediction: Periaxin anchors the Schwann cell to its basal lamina through the DRP2 and dystroglycan complex that shapes Cajal bands, and the literature predicts that truncating PRX variants on both alleles simply take it away. Prx-null mice lose those appositions, myelinate over abnormally short internodes, and develop demyelination with marked sensory abnormality, matching the prominent sensory loss that sets affected patients apart from typical CMT1.

Rationale: The lesion here is architectural: with no periaxin the abaxonal cytoplasmic channels collapse, internodal length is set wrong from the start, and sensory fibers suffer disproportionately. Nonsense and frameshift alleles predominate, so no mutant periaxin persists in the sheath to act on anything.

Learn more about CMT4F →

Mechanistic basis: Hypomorphic

Prediction: Biochemical evidence supports a loss-of-function mechanism for CMT-PSAT1: phosphoserine aminotransferase catalyzes the second step of de novo L-serine synthesis, converting 3-phosphohydroxypyruvate to phosphoserine, and biallelic variants lower that activity enough to limit serine and downstream glycine supply to nerve. Patients show reduced serine in plasma and cerebrospinal fluid, and oral serine raises those levels, a direct demonstration that what is missing is enzymatic output.

Rationale: The serine pathway is a supply chain and the lesion sits at one enzymatic step within it: reduced flux, with no toxic protein species involved. Response to serine, which bypasses the block entirely, is the strongest evidence available for that reading. CMT sits at the mild end of a spectrum whose severe end is Neu-Laxova syndrome, and that phenotypic breadth is what leaves confidence at medium.

Learn more about CMT-PSAT1 →

Mechanistic basis: Overactivity

Prediction: The literature predicts a gain-of-function mechanism for CMT2B: dominant RAB7A missense mutations such as L129F, K157N and V162M ring the nucleotide-binding pocket of the Rab7 GTPase and accelerate GDP/GTP exchange, leaving the protein in its active GTP-bound state far longer than normal. The constitutively active GTPase perturbs late endosomal and lysosomal trafficking and NGF/TrkA signaling in sensory neurons, which fits the striking sensory predominance of the phenotype.

Rationale: Every mapped CMT2B substitution sits at the nucleotide-binding site and produces the same biochemical outcome, a Rab7 that cycles too fast and stays switched on, so the mutant is doing more of Rab7's own work rather than obstructing the normal protein. That is overactivity, and one such allele is sufficient.

Learn more about CMT2B →

Mechanistic basis: Haploinsufficiency

Prediction: Evidence across REEP1 families predicts a haploinsufficiency mechanism for dHMN-5B: most dominant alleles are truncating variants or whole-exon deletions whose transcripts are cleared by nonsense-mediated decay, leaving a single working copy of a dosage-sensitive ER-shaping and microtubule-interacting protein, and Reep1-null models show the expected axonal degeneration. Loss of the same gene also causes SPG31, a dominant hereditary spastic paraplegia, so half-normal REEP1 is on its own enough to injure long axons.

Rationale: Nonsense-mediated decay of the common truncating alleles removes protein instead of producing a mutant that could act on its partner, and tubular ER morphology depends on REEP1 concentration, which is why one intact copy falls short. C-terminal variants that escape decay may add a toxic component, and that open question leaves the mechanism graded medium.

Learn more about dHMN-5B →

Mechanistic basis: Complete loss

Prediction: Loss of an ER-phagy receptor is what the literature predicts for HSAN-2B: recessive nonsense and frameshift alleles in RETREG1 (FAM134B) remove the reticulophagy receptor that shapes endoplasmic reticulum membranes and delivers ER fragments to autophagic turnover. Sensory neurons, with their long axons and heavy secretory load, degenerate once both copies are inactivated, and supplying the intact receptor is predicted to restore ER turnover.

Rationale: The alleles are truncating and the phenotype needs two of them, which is the signature of a missing receptor rather than a mutant one acting on its partners. RETREG1 depletion impairs ER membrane remodeling and reticulophagy in cell models, tying the genetic pattern to a defined cell-biological deficit in long sensory axons.

Learn more about HSAN-2B →

Mechanistic basis: Hypomorphic

Prediction: The literature predicts a biallelic loss-of-function mechanism for CMT-RFC1: the disease allele is an intronic pentanucleotide expansion, most often AAGGG, replacing the reference AAAAG repeat inside an Alu element of RFC1, which encodes the large subunit of the replication factor C clamp loader. Expanded repeats on both copies reduce functional RFC1 and produce the sensory neuronopathy of CANVAS. Although the lesion is an expansion, the supported mechanism is loss of the clamp loader's normal activity, not a toxic repeat product.

Rationale: Expansions carry an automatic suspicion of toxic gain, and this locus argues against it: truncating RFC1 point variants found in trans with a single expansion produce the same disease, which is what a loss-of-function allelic series looks like. Whether the expansion acts purely by reducing expression or also through a repeat-derived species is still argued, and that open question holds the grade at medium.

Learn more about CMT-RFC1 →

Mechanistic basis: Complete loss

Prediction: Published work on ARSACS supports a biallelic loss-of-function mechanism for CMT-SACS: SACS encodes sacsin, a 4,579-residue chaperone built from a ubiquitin-like domain, HSP90-like ATPase repeat regions, a DnaJ domain that recruits HSP70, and a HEPN domain. Recessive truncating variants, including the Quebec founder alleles c.6594delT and c.5254C>T, remove the protein, while destabilizing missense alleles deplete it. Cells without sacsin show bundled neurofilaments and abnormally fused, mislocalized mitochondria, the signature seen in patient fibroblasts and Sacs-null mice.

Rationale: Sacsin behaves as a dose-dependent chaperone: obligate carriers of the founder truncations are neurologically normal, while cells with none of the protein show the full neurofilament and mitochondrial phenotype. Nonsense, frameshift, splice, and destabilizing missense alleles all converge on one clinical picture, the allelic pattern of a deficiency and not of a mutant species acting on its wild-type partner.

Learn more about CMT-SACS →

Mechanistic basis: Unresolved

Prediction: The evidence does not converge on a mechanism for CMT-SARS1: this ultra-rare dominant subtype is known from a single de novo report (Record et al. 2023) with no functional dissection, and the mutant seryl-tRNA synthetase has been assayed neither for aminoacylation nor for the conformational opening that generates neomorphic interactions in other dominant tRNA-synthetase CMTs. Lost serylation capacity and a toxic gain acquired by the mutant enzyme both remain in play.

Rationale: Mechanism is heterogeneous across the dominant tRNA-synthetase CMTs, interference in some and neomorphic activity in others, so what holds for one gene does not transfer to SARS1. Absent aminoacylation, supplementation, or model-organism data on the de novo alleles, the mechanism stays open.

Learn more about CMT-SARS1 →

Mechanistic basis: Complete loss

Prediction: Evidence in CMT4B3 supports the same pathway lesion one gene over: SBF1 encodes MTMR5, the pseudophosphatase paralogue of MTMR13, and the recessive variants reported are predicted to strip its scaffolding of the MTMR2 phosphatase complex, with two defective copies required for disease. Reported families range from a severe childhood demyelinating course to a syndromic presentation, and focally folded myelin is not uniformly present.

Rationale: Scaffolding is the whole of MTMR5's contribution, so a recessive SBF1 defect leaves the MTMR2 complex unassembled where the myelin membrane needs it. Confidence stops short of the CMT4B1 level for a specific reason: several reported SBF1 changes are missense of undetermined functional consequence, and the clinical picture spans more than one syndrome.

Learn more about CMT4B3 →

Mechanistic basis: Complete loss

Prediction: Truncating and domain-disrupting SBF2 variants remove MTMR13, a catalytically dead myotubularin that binds MTMR2 and raises its phosphatase output, and the literature predicts that CMT4B2 is the loss of that partner protein across both alleles. Nerve biopsy shows the focally folded myelin also seen in CMT4B1, placing the two diseases on one pathway, and some reported families additionally segregate early-onset glaucoma.

Rationale: Because MTMR13 has no catalytic site of its own, its only route to disease is failure to activate MTMR2, and the shared outfolding pathology confirms that the two genes act in series. An allele spectrum dominated by premature stops leaves no partial product in the complex.

Learn more about CMT4B2 →

Mechanistic basis: Overactivity

Prediction: The literature predicts a gain of function in HSAN-7 at the level of a single channel: heterozygous de novo SCN11A missense variants, classically p.Leu811Pro, shift Nav1.9 toward a persistently open state, and the resulting standing inward sodium current holds nociceptors chronically depolarized. Sodium channels needed for the action potential upstroke sit inactivated at that membrane potential, so the neuron cannot fire, and the clinical result is congenital insensitivity to pain with self-mutilation and slow-healing injuries. Variants that lower Nav1.9 activity instead track with painful small-fiber phenotypes.

Rationale: Persistent current, not reduced excitability, is the primary lesion, and the paradox of a hyperactive channel silencing its neuron resolves once depolarization block is taken into account. The mutant channel is overactive in heterologous expression on its own terms, with no requirement that it act on the wild-type allele, and the opposite pain phenotype at the same gene shows that direction of effect matters more than dose.

Learn more about HSAN-7 →

Mechanistic basis: Complete loss

Prediction: Two inactive copies of a single sodium channel gene account for HSAN-2D, and the evidence predicts biallelic loss of function: nonsense, frameshift, and splice alleles in SCN9A eliminate Nav1.7 from nociceptors and sympathetic ganglion neurons, removing the amplifier that brings small-fiber terminals to action potential threshold. Anosmia accompanies the pain insensitivity because the same channel is required in olfactory sensory neurons. Wild-type Nav1.7 add-back is predicted to rescue.

Rationale: Direction of effect separates the SCN9A diseases cleanly: channel-inactivating alleles in two copies abolish pain sensation from birth, while heterozygous alleles that slow inactivation or shift activation produce erythromelalgia and paroxysmal extreme pain disorder. HSAN-2D falls on the inactivating side, and relatives carrying a single null allele have normal pain perception.

Learn more about HSAN-2D →

Mechanistic basis: Hypomorphic

Prediction: The literature predicts a biallelic loss-of-function mechanism for CMT-SCO2: SCO2 is a copper metallochaperone that delivers copper to the CuA site of cytochrome c oxidase subunit II, without which complex IV cannot be assembled. The recurrent hypomorphic p.Glu140Lys allele, typically carried in trans with a truncating variant, leaves partial metallation capacity, and cytochrome c oxidase activity is reduced in patient tissue. Restoring the chaperone is predicted to rescue complex IV assembly.

Rationale: Copper delivery is catalytic and quantitative, so severity scales with how much metallation capacity survives: near-complete inactivation of both copies gives fatal infantile cardioencephalomyopathy, while p.Glu140Lys-containing genotypes retain enough function for a later-onset presentation dominated by peripheral nerve. That graded genotype-phenotype relation is the argument for residual-activity loss over interference. Few families carry the CMT presentation, hence a medium grade.

Learn more about CMT-SCO2 →

Mechanistic basis: Complete loss

Prediction: Human null alleles and a long-studied mouse converge on biallelic loss of function for CMT-SCYL1: SCYL1 encodes a catalytically inactive pseudokinase that binds COPI coat components and supports retrograde Golgi-to-endoplasmic-reticulum traffic. Recessive nonsense, frameshift, and splice variants remove it, and the spontaneously arising Scyl1-deficient mdf mouse develops motor neuron degeneration from the same absence of protein. Affected individuals present with the CALFAN combination of cholestatic liver episodes, ataxia, and CMT.

Rationale: Truncating alleles spread across the gene produce one phenotype, and the mdf mouse arrives at it by losing the protein entirely, so disease tracks absent SCYL1 rather than any property of a mutant version. Motor neurons are the vulnerable cell type because retrograde traffic sustains the longest secretory routes in the body.

Learn more about CMT-SCYL1 →

Mechanistic basis: Dosage

Prediction: CMT-SEPTIN9 is predicted in the literature to act through gene dosage: intragenic duplications within SEPTIN9 raise the amount of a normal-sequence septin rather than removing a working copy, so the pathogenic quantity is an excess and supplementation is not predicted to help. Clustered N-terminal missense alleles are also reported and would work differently, by disrupting assembly of the hetero-oligomeric septin filament, and that unresolved split between two allele classes sets confidence at medium.

Rationale: An extra normal-sequence copy cannot cause disease through insufficiency, which is the argument that separates the duplication alleles from haploinsufficiency and places them with excess septin. The N-terminal missense cluster points elsewhere, toward interference within the filament, and the two allele classes have not been reconciled.

Learn more about CMT-SEPTIN9 →

Mechanistic basis: Complete loss

Prediction: The literature predicts a biallelic loss-of-function mechanism for CMT-SETX: senataxin is an RNA/DNA helicase that resolves R-loops and supports transcription termination, and two null or hypomorphic SETX alleles drop that activity below the threshold neurons tolerate, giving ataxia with oculomotor apraxia type 2 with elevated alpha-fetoprotein and sensorimotor CMT. Heterozygous SETX missense variants such as p.Leu389Ser instead cause ALS4, a dominant gain-of-function-like disease at the same locus.

Rationale: Senataxin activity is threshold-dependent, and the two SETX diseases separate cleanly on that axis: nonsense, frameshift, and splice alleles must be present on both copies to cause AOA2, while single missense alleles in ALS4 act through something the null alleles never reproduce. How R-loop accumulation connects reduced helicase activity to the neuronal phenotype is not fully mapped, and the medium grade reflects that gap.

Learn more about CMT-SETX →

Mechanistic basis: Hypomorphic

Prediction: Enzymatic evidence supports biallelic loss of function in CMT-SGPL1: sphingosine-1-phosphate lyase catalyzes the only irreversible exit from sphingolipid metabolism, cleaving sphingosine-1-phosphate into hexadecenal and phosphoethanolamine. Compound heterozygous and homozygous null or hypomorphic alleles leave patient cells with measurably reduced lyase activity and accumulated sphingoid base phosphates, producing S1P lyase insufficiency syndrome with steroid-resistant nephrotic syndrome, adrenal insufficiency, and peripheral nerve involvement.

Rationale: Measured activity in patient cells, not inference from variant type, anchors this one: lyase activity falls with the alleles carried, and heterozygous parents retain enough of it to stay well. Substrate build-up sits downstream of the enzymatic block rather than being an independent toxic property of the variant protein, so replacing lyase activity addresses both.

Learn more about CMT-SGPL1 →

Mechanistic basis: Complete loss

Prediction: The literature predicts a biallelic loss-of-function mechanism for CMT4C: the recurrent p.Arg954* allele, together with the frameshift, splice, and trafficking-disrupting missense variants that accompany it, leaves Schwann cells without SH3TC2 at the recycling endosome, where the protein delivers membrane to the growing sheath and to the node. Carriers are unaffected, Sh3tc2-null mice reproduce the nodal and myelination defect, and an intact copy of the gene is predicted to restore that delivery.

Rationale: The early scoliosis that marks this subtype follows from myelination failing while the nerve is still growing rather than from a later toxic process, and p.Arg954*, common in European cohorts, yields no stable protein that could act on the wild type. Homozygotes and compound heterozygotes for two loss alleles are affected alike.

Learn more about CMT4C →

Mechanistic basis: Complete loss

Prediction: In the reported SIGMAR1 families, the literature predicts biallelic loss of function: homozygous variants destabilize the sigma-1 receptor, a chaperone anchored at mitochondria-associated ER membranes where it governs calcium transfer to mitochondria and the folding of client proteins, so motor neurons lose that support only when both copies are affected. Sigmar1-null mice develop progressive motor weakness, matching what full absence of the receptor produces.

Rationale: Destabilized receptor is degraded rather than retained in a form able to act on its wild-type counterpart, and null mice reproduce the motor phenotype. Whether the aggregates formed by some mutant receptors add toxicity of their own, or merely mark the disposal of an unstable protein, is unsettled; medium confidence tracks that question.

Learn more about dHMN2-SIGMAR1 →

Mechanistic basis: Overactivity

Prediction: The evidence predicts an overactivity gain-of-function mechanism for CMT2II: the de novo SLC12A6 variant Thr991Ala and related dominant alleles remove the inhibitory C-terminal phosphoregulation of KCC3, locking the K-Cl cotransporter in a constitutively active state and driving unregulated potassium chloride efflux with the cell volume consequences that follow in peripheral nerve. Biallelic loss of SLC12A6 produces a different disease entirely, the agenesis of the corpus callosum with peripheral neuropathy of Andermann syndrome, which marks the dominant alleles as active rather than inactive transporters.

Rationale: Substitutions at the regulatory C-terminus mimic the dephosphorylated state and leave KCC3 permanently switched on, and a transporter running without its brake does damage regardless of how much normally regulated KCC3 sits beside it. That two inactivating alleles are required for Andermann syndrome sharpens the distinction. Only a handful of dominant alleles have been characterized functionally, which is the limit on confidence here.

Learn more about CMT2II →

Mechanistic basis: Hypomorphic

Prediction: Protein-level data predict biallelic loss of function in HMSN-6B: SLC25A46 is an outer mitochondrial membrane protein related to the fusion machinery, working alongside MFN2 and the MICOS complex to balance mitochondrial fission against fusion and to maintain membrane contacts. Pathogenic missense variants destabilize the protein and truncating alleles remove it, and how much protein remains at steady state predicts where a family sits on a spectrum running from CMT with optic atrophy to lethal congenital pontocerebellar hypoplasia.

Rationale: Steady-state abundance is the variable that matters: mutant SLC25A46 is degraded rapidly, and the severity gradient across families follows how much survives rather than which residue changed. Optic atrophy joins the peripheral disease because retinal ganglion cells share the dependence on mitochondrial dynamics that MFN2 and OPA1 disease demonstrates.

Learn more about HMSN-6B →

Prediction: The literature predicts a dominant-negative mechanism for dHMN-7A: SLC5A7 encodes CHT1, the high-affinity choline transporter whose delivery to the presynaptic surface sets the rate of acetylcholine resynthesis, and the heterozygous C-terminal truncating alleles found with this distal motor phenotype and vocal cord paralysis strip trafficking determinants from the tail, leaving a mislocalized transporter that pulls choline uptake below what one intact allele sustains. Biallelic null SLC5A7 alleles instead cause a recessive congenital myasthenic syndrome.

Rationale: Uptake in these heterozygotes falls further than a halved dose explains, and that shortfall is what separates interference from simple loss: truncated CHT1 lacking its trafficking signals holds normal transporter away from the terminal membrane. Carriers of a single null allele stay well, so one working copy suffices. Evidence that the mutant acts on the normal transporter, rather than only mistrafficking itself, remains indirect.

Learn more about dHMN-7A →

Mechanistic basis: Complete loss

Prediction: Genetic, biochemical, and therapeutic evidence converge on biallelic loss of function for CMT-SORD: sorbitol dehydrogenase catalyzes the second step of the polyol pathway, oxidizing sorbitol to fructose, and the recurrent frameshift c.757delG, which lies in the segment duplicated between SORD and the SORD2P pseudogene and was long missed by standard sequencing, abolishes enzyme activity. Two defective copies produce marked serum and nerve sorbitol elevation, while heterozygous carriers, who are numerous, remain unaffected.

Rationale: Sorbitol accumulation is the measurable consequence of the missing enzyme, and lowering it upstream with aldose reductase inhibitors, now in clinical development for this disease, reduces the metabolite in treated patients: therapeutic confirmation that the pathology follows from absent sorbitol dehydrogenase activity and not from anything a truncated product does. Carrier status without phenotype completes the dosage argument.

Learn more about CMT-SORD →

Mechanistic basis: Complete loss

Prediction: Spatacsin partners with spastizin in autophagic lysosome reformation, and the literature predicts loss of that activity on both SPG11 alleles in CMT2X: the variants reported in this predominantly motor recessive phenotype are truncating, removing the protein rather than yielding an altered one. Heterozygous relatives carrying a single truncating allele are neurologically normal, and no CMT2X variant on record produces a stable mutant spatacsin.

Rationale: Truncating SPG11 alleles delete spatacsin from the lysosome-reformation machinery, and nothing in the reported allele spectrum implicates a residual product acting on the wild type. What stays unsettled is why the same null genotypes yield CMT2X in some families and spastic paraplegia with a thin corpus callosum or juvenile ALS in others, so confidence sits at medium.

Learn more about CMT2X →

Mechanistic basis: Neomorphic

Prediction: For HSAN-1A the literature predicts a neomorphic gain of function: serine palmitoyltransferase normally condenses palmitoyl-CoA with L-serine, and the dominant SPTLC1 missense alleles relax that specificity so the enzyme also accepts L-alanine and glycine. The products are 1-deoxysphingolipids, atypical bases the mutant enzyme makes and the normal enzyme does not, and they accumulate in plasma and in sensory neurons. Separately, SPTLC1 alleles that escape ORMDL3-mediated feedback drive unrestrained canonical SPT activity and childhood-onset ALS.

Rationale: Swapping alanine or glycine into the condensation step changes what the enzyme makes, not how much of it there is. Canonical sphingolipid output is largely maintained while a lipid species healthy tissue barely carries appears at high levels, which is the signature of a new activity rather than a missing one. Nothing in that chemistry requires the wild-type subunit to be affected.

Learn more about HSAN-1A →

Mechanistic basis: Neomorphic

Prediction: Two recurrent SPTLC1 alleles anchor the prediction of a neomorphic gain of function in HSN-1A: C133W and V144D replace residues lining the substrate channel at the interface of the SPTLC1/SPTLC2 heterodimer, and the remodeled pocket admits alanine and glycine alongside serine, yielding neurotoxic 1-deoxysphingolipids. Sptlc1 knockout does not reproduce the sensory phenotype, and added wild-type enzyme fails to neutralize the atypical lipids, which the literature reads as an acquired activity rather than lost or disrupted function.

Rationale: Neither test that would expose a loss comes back positive: removing Sptlc1 does not recreate the disease, and restoring normal enzyme leaves the 1-deoxysphingolipid burden in place. Both results point away from haploinsufficiency and away from interference with the partner subunit, and toward a mutant enzyme that manufactures something new by itself.

Learn more about HSN-1A →

Mechanistic basis: Neomorphic

Prediction: Across HSAN-1C the reported evidence predicts a gain of new activity rather than loss of the old: dominant SPTLC2 missense alleles redirect serine palmitoyltransferase onto alanine and glycine, and the resulting 1-deoxysphingolipids lack the C1 hydroxyl group that normal sphingoid bases carry. Without it they can neither be built into complex sphingolipids nor cleared through the canonical degradation route, so they persist, and sensory neurons exposed to them retract and lose neurites. The toxicity belongs to the metabolite, not to any shortfall in SPT activity.

Rationale: Because 1-deoxysphingolipids cannot enter the normal catabolic pathway, dorsal root ganglion neurons carry a lipid they have no route to dispose of, and the longest sensory axons show it first. Supplying more normal enzyme adds canonical product but removes none of the atypical one, which is why the evidence reads as acquired toxicity rather than haploinsufficiency.

Learn more about HSAN-1C →

Mechanistic basis: Neomorphic

Prediction: In HSN-1C the literature places the neomorphic gain in the catalytic half of the enzyme: SPTLC2 carries the pyridoxal-5'-phosphate-binding lysine of the serine palmitoyltransferase heterodimer, and dominant missense substitutions there widen the amino acid preference of the active site to include alanine and glycine. The condensation reaction then yields 1-deoxysphingolipids in place of canonical sphingoid bases. Because SPTLC1 contributes the non-catalytic subunit of the same enzyme, alleles in either gene converge on one metabolic lesion.

Rationale: SPTLC2 substitutions act on the chemistry itself rather than on an accessory subunit, which is why the metabolic readout is so clean. Excess L-serine lowers circulating 1-deoxysphingolipids in this disease group by outcompeting alanine at a mutant active site, and that response is diagnostic: an enzyme that had simply stopped working would generate no such metabolite to lower.

Learn more about HSN-1C →

Mechanistic basis: Complete loss

Prediction: The literature predicts a biallelic loss-of-function mechanism for CMT4K: SURF1 encodes a nuclear-encoded assembly factor for cytochrome c oxidase, implicated in heme a delivery during the early steps of complex IV biogenesis. The index family is homozygous for the splice-acceptor allele c.107-2A>G, which yields no normally spliced transcript and no detectable SURF1 protein. Carriers of one such allele are unaffected, and restoring functional SURF1 is predicted to restore assembly.

Rationale: SURF1 is absent in these patients, and what remains partial is complex IV downstream, assembled inefficiently without the factor rather than not at all. Surf1-null mice are viable, so losing the factor outright is survivable, and the deficiency patient tissue shows is quantitative rather than a novel toxic species. What varies across SURF1 disease is which system declines first, not how the alleles act.

Learn more about CMT4K →

Prediction: The literature predicts a dominant-negative mechanism for CMT-SYT2: heterozygous C2B-domain missense variants in SYT2 (for example p.Pro308Leu, p.Asp307Ala) produce a mutant synaptotagmin-2 that incorporates into the release machinery and interferes with calcium-triggered synaptic vesicle exocytosis, impairing presynaptic neurotransmission. Because the mutant sensor is built into the release machinery, added wild-type protein is not predicted to rescue. Biallelic null alleles instead cause a separate recessive presynaptic myasthenic syndrome, a loss-of-function phenotype distinct from this dominant missense mechanism.

Rationale: Mutant synaptotagmin-2 from the heterozygous C2B-domain missense variants incorporates into the release machinery and disrupts calcium-triggered vesicle exocytosis, which points to a dominant-negative effect rather than loss of function. That biallelic nulls instead cause a separate recessive presynaptic myasthenic syndrome supports the distinction, and because the mutant sensor is built into the release apparatus, wild-type add-back is not predicted to rescue.

Learn more about CMT-SYT2 →

Mechanistic basis: Complete loss

Prediction: Autophagosome formation fails in HSAN-9, and the evidence predicts a biallelic loss-of-function mechanism: recessive TECPR2 variants are predominantly truncating, including a founder frameshift in Bukharian Jewish families, and they eliminate an autophagy adaptor that also supports ER-to-Golgi trafficking. Sensory and autonomic degeneration appears alongside central hypoventilation and developmental involvement, consistent with the breadth of the protein's role. Wild-type TECPR2 add-back is predicted to rescue.

Rationale: Truncating alleles on both chromosomes with unaffected heterozygous parents point to an absent adaptor, and patient-derived cells show impaired autophagosome formation, linking genotype to pathway directly. The wide phenotype follows from TECPR2 acting in general autophagic flux rather than in a sensory-specific program.

Learn more about HSAN-9 →

Mechanistic basis: Neomorphic

Prediction: Evidence in HMSN-Okinawa Type predicts a toxic gain of function: the recurrent heterozygous TFG p.Pro285Leu substitution sits in the C-terminal low-complexity region and raises the protein's tendency to self-assemble into insoluble cytoplasmic inclusions, with proteasomal handling compromised in affected motor neurons. Biallelic TFG variants instead cause a recessive spastic paraplegia with optic atrophy, so reduced TFG dosage does not account for the proximal-dominant phenotype.

Rationale: Insoluble inclusions, not diminished ER-exit-site function, dominate the pathology, and the mutant acquires that aggregation property on its own. A separate recessive TFG disease is the strongest argument that losing this protein's activity produces something else entirely. What keeps the mechanism unsettled is TFG's native oligomerization: an aggregating mutant can plausibly draw wild-type subunits in with it.

Learn more about HMSN-Okinawa Type →

Mechanistic basis: Complete loss

Prediction: Truncating TRIM2 alleles on both chromosomes, homozygous or compound heterozygous, abolish an E3 ubiquitin ligase whose substrate is neurofilament light chain. On that basis the evidence supports a recessive loss of function in CMT2R, with unubiquitinated neurofilament accumulating in axons; Trim2-null mice develop a comparable early-onset axonal degeneration. None of the reported alleles encodes a product capable of competing with a functioning ligase.

Rationale: Accumulation of neurofilament light chain follows directly from removing the ligase that clears it, and truncating alleles on both chromosomes leave no residual enzyme to perform that turnover. In these homozygous families there is no wild-type ligase present for a mutant one to act against, which settles the mechanism as a straightforward recessive loss.

Learn more about CMT2R →

Mechanistic basis: Overactivity

Prediction: The evidence supports a gain-of-function mechanism for CMT2C: dominant TRPV4 missense mutations such as R269H, R315W and R316C cluster on one face of the intracellular ankyrin-repeat domain and produce a calcium channel with elevated basal open probability, so cells carrying a single mutant allele take up calcium at rest. Patch-clamp recordings of these channels show increased rather than diminished current, and the resulting chronic calcium load is toxic to motor and sensory neurons.

Rationale: A channel that opens when it should be shut is doing its own work in excess, not obstructing the wild-type subunits it assembles with. The homozygous p.Ser94Leu allele fits the same mechanism at lower amplitude: it reaches the membrane normally, scores as a moderate rather than strong overactive channel, and its toxicity is reversed by a TRPV4 antagonist, so two copies are needed to cross threshold where the ankyrin-repeat alleles cross it with one.

Learn more about CMT2C →

Mechanistic basis: Overactivity

Prediction: Experimental evidence supports a gain-of-function mechanism for dHMN-8: dominant TRPV4 missense variants cluster in the cytoplasmic ankyrin-repeat domains, and the substituted channels sit open at rest, driving basal calcium influx to levels motor neurons do not survive. Cell-based recordings show elevated constitutive current rather than reduced conductance, Trpv4-null animals do not reproduce the scapuloperoneal and distal weakness pattern, and it is channel blockade, not added wild-type protein, that corrects the cellular defect.

Rationale: Calcium overload through a channel that never fully closes is the proximate injury, which is why the phenotype tracks with the ankyrin-repeat substitutions that raise basal current. Absence of TRPV4 produces nothing resembling dHMN-8, which rules out a dosage mechanism, and antagonists that shut the channel are what reverse the toxicity in cellular models.

Learn more about dHMN-8 →

Prediction: The evidence predicts a dominant-negative mechanism for CMT-TUBB3: tubulin functions only as a polymer, so heterozygous TUBB3 missense variants place mutant beta-tubulin into alpha/beta-heterodimers and then into microtubules copolymerized with wild-type subunits, altering lattice dynamics and kinesin interactions and impairing axonal transport. No TUBB3 disease arises from deletion or truncation, and no haploinsufficiency phenotype has been described, which leaves interference inside the shared polymer as the mechanism these alleles support.

Rationale: Because every microtubule is built from both alleles, a mutant subunit changes the behavior of a structure the wild-type protein has to share, and added wild-type is not predicted to restore a lattice that keeps incorporating it. Some groups read the altered kinesin interactions as a toxic gain of function instead, and that unresolved reading sets the grade at medium.

Learn more about CMT-TUBB3 →

Mechanistic basis: Hypomorphic

Prediction: The literature predicts a loss-of-function mechanism for dSMAX-2: UBA1 encodes the sole E1 enzyme that activates ubiquitin and charges the downstream E2 cascade, and the hypomorphic exon 15 missense alleles (p.Met539Ile, p.Ser547Gly) together with expression-lowering regulatory variants leave residual activity too low to sustain protein turnover in motor neurons. Hemizygous males carry that reduced-function copy as their only UBA1, and restored wild-type E1 is predicted to rescue. Somatic p.Met41 variants confined to the myeloid lineage instead cause VEXAS syndrome, an acquired inflammatory disease.

Rationale: A hemizygous male expresses no wild-type E1 for a mutant subunit to interfere with, so interference is unavailable as a mechanism and the phenotype tracks residual enzyme activity alone. Because ubiquitin activation is essential to every cell, only partial-function alleles are compatible with survival, and the motor neuron is where falling conjugation capacity shows first.

Learn more about dSMAX-2 →

Mechanistic basis: Unresolved

Prediction: The literature does not resolve a mechanism for dHMN1-UBE3C. Cutrupi et al. (2023) traced the disease to a genomic insertion creating a UBE3C-intergenic fusion transcript, which splices duplicated UBE3C exons 1 to 10 onto a pseudo-exon. The transcript escapes nonsense-mediated decay, its product stops short of the HECT catalytic domain, and full-length UBE3C is reduced in patient motor neurons, but ubiquitin ligase activity itself was never measured.

Rationale: Three readings fit the same observation and predict different mechanisms: the truncated product acting on the normal protein, transcriptional interference in cis at the duplicated locus, or a toxicity belonging to the fusion product, which is what expressing it in neurons produces. What the evidence does settle is the negative case, since biallelic UBE3C loss causes a separate recessive neurodevelopmental disease and its heterozygous carriers have no neuropathy, ruling out a simple halved dose.

Learn more about dHMN1-UBE3C →

Mechanistic basis: Neomorphic

Prediction: The literature predicts a neomorphic gain-of-function mechanism for CMT2Y: dominant VCP missense substitutions cluster at the N-domain/D1 interface of p97, where they loosen the interdomain communication that normally restrains the enzyme, producing elevated ATPase activity and altered cofactor recruitment. Autophagic and proteasomal clearance falter downstream, and TDP-43 mislocalizes and aggregates. Reduced p97 dose does not produce disease, so the dominant phenotype cannot be read as haploinsufficiency.

Rationale: An enzyme hydrolyzing ATP faster and handing substrates to the wrong cofactors is doing too much of its own job, not too little, and that surplus is what derails clearance of ubiquitinated and TDP-43 substrates. The evidence comes largely from multisystem proteinopathy families carrying the same residues, and because p97 assembles as a hexamer from both alleles, a disrupted-subunit contribution stays open at medium confidence.

Learn more about CMT2Y →

Mechanistic basis: Hypomorphic

Prediction: Kinase activity is the graded variable in VRK1-related dSMA, and the literature predicts loss of function across both alleles: compound-heterozygous or homozygous variants lower VRK1 serine/threonine kinase output, which motor neurons need for Cajal body assembly, nuclear envelope dynamics, and DNA damage repair. Combinations retaining partial activity present as distal motor disease, while more complete loss of the same enzyme causes pontocerebellar hypoplasia type 1A, a severity range that follows enzyme dose.

Rationale: One gene spans a lethal pontocerebellar disorder and a distal motor phenotype, and the ordering variable is how much kinase activity a genotype retains. A dominant-negative mutant would not generate that graded series, and heterozygous parents throughout the reported families are unaffected.

Learn more about dSMA →

Mechanistic basis: Complete loss

Prediction: The literature predicts a biallelic loss-of-function mechanism for dHMN-VWA1: the recurrent founder frameshift c.62_71dup and other truncating alleles eliminate WARP, a small secreted von Willebrand factor A domain protein that resides in the basement membrane of peripheral nerve and at the neuromuscular junction. Two damaged alleles are required, unaffected heterozygous parents are the rule in reported pedigrees, and Vwa1-null mice show the matching nerve and junctional abnormalities.

Rationale: WARP is secreted, so a frameshift allele yielding no protein leaves nothing in the matrix that could act on the product of the other copy. Heterozygous carriers of the founder duplication are unaffected across the published families, placing the disease threshold at near-complete depletion of matrix WARP.

Learn more about dHMN-VWA1 →

Prediction: The literature predicts a dominant-negative mechanism for dHMN-9: tryptophanyl-tRNA synthetase functions as a homodimer, and the heterozygous WARS1 substitutions reported in this dominant distal motor phenotype lower aminoacylation while remaining competent to pair with normal subunit, so charged tRNA-Trp falls further in long motor axons than a single intact allele would allow. Biallelic WARS1 loss produces a distinct multisystem neurodevelopmental disorder rather than selective distal weakness, and with no in vivo model of the dominant alleles the mechanism stays debated, so confidence is medium.

Rationale: Reduced tryptophan charging alone does not account for a dominant phenotype, and that gap is the signature of a mutant subunit disrupting its dimer partner rather than acting as an independent gain of function. The interference is read from WARS1 biochemistry, which is why it has not been settled against the neomorphic models proposed for other dominant tRNA synthetase alleles.

Learn more about dHMN-9 →

Mechanistic basis: Complete loss

Prediction: The literature predicts a biallelic loss-of-function mechanism for HSAN-2A: recessive truncating variants cluster in HSN2, the nervous-system-specific exon of WNK1, and remove the neuronal isoform while leaving the ubiquitously expressed kinase intact. Both copies of that isoform must be lost before disease appears. Restored wild-type is predicted to rescue. Separate WNK1 alleles that raise kinase signaling cause pseudohypoaldosteronism type II, a hypertension disorder, so the two directions of effect at this locus give unrelated diseases.

Rationale: Because the truncations fall in an exon used only by the neuronal transcript, the deficit is confined to the WNK1/HSN2 isoform, which explains why single-allele carriers are unaffected and why the kinase's renal functions are spared. Gain-of-function WNK1 alleles at the same locus drive hypertension instead, making direction of effect rather than the gene the determinant of phenotype.

Learn more about HSAN-2A →

Mechanistic basis: Complete loss

Prediction: Nonsense-mediated decay is the operative lesion in HSN-2A, and the literature predicts biallelic loss of function: premature stop codons in the HSN2 exon send the neuron-specific WNK1 transcript to degradation, and alleles that escape decay yield a truncated product stripped of the kinase context it serves. Disease requires both copies of the neuronal isoform to be inactivated, and returning that isoform is predicted to restore function in sensory neurons.

Rationale: Transcript degradation, not a toxic peptide, is what the HSN2 stop codons produce, so the neuronal WNK1 pathway is simply absent in homozygotes and compound heterozygotes while single-allele carriers retain normal sensation. Restriction of HSN2 expression to the nervous system confines that absence to sensory and autonomic neurons.

Learn more about HSN-2A →

Mechanistic basis: Neomorphic

Prediction: The evidence supports a neomorphic gain-of-function mechanism for CMTDIC: the dominant YARS1 alleles G41R, E196K, and the 153-156 VKQV deletion loosen the tyrosyl-tRNA synthetase into a conformation that mislocalizes within axons and acquires binding partners the normal enzyme does not engage, as demonstrated in Drosophila and yeast. Because reduced aminoacylation is necessary but not sufficient for the phenotype in those models, and because the toxicity travels with the mutant protein, the evidence favors a new activity over a diminished enzyme supply.

Rationale: Toxicity here is dissociable from catalysis, since the mutant enzyme damages axons in fly and yeast models in ways that reduced aminoacylation alone does not account for, and added wild-type YARS1 does not correct the phenotype. Which partners the mislocalized enzyme engages, and how much the residual catalytic deficit contributes, both remain open, so the grade is medium.

Learn more about CMTDIC →

Mechanistic basis: Gene unknown

Prediction: For CMTX2 the evidence stops at a locus: linkage in the historical pedigrees places the disease at Xp22.2, an interval well separated from GJB1 at Xq13.1, making this a genetically distinct X-linked entity rather than an allelic variant of CMTX1. Because the transcript within that interval has never been identified, there is no protein product and no allele class, and no mechanism can be specified.

Rationale: Mapping data are the entire evidence base here, and a chromosomal position establishes where the defect lies without saying what it is, which leaves no substrate for inference about loss, interference, or gain. The separation from GJB1 does at least exclude the connexin-32 biology that accounts for CMTX1.

Learn more about CMTX2 →

Mechanistic basis: Gene unknown

Prediction: Harding's classification defined dHMN-1 by juvenile onset and dominant transmission of distal motor weakness without sensory loss, criteria that group families rather than identify a locus. No gene has been definitively established for the type and no linkage interval fixes it, leaving not even a candidate genomic address from which the literature could reason toward a mechanism.

Rationale: Clinical criteria assembled this type, and clinical criteria cannot yield a molecular lesion. Distal motor neuropathy genes found since, acting through varied mechanisms, carry their own designations, so dHMN-1 has neither a protein nor an allele class whose behavior could be predicted.

Learn more about dHMN-1 →

Mechanistic basis: Gene unknown

Prediction: HMSN-5 is a phenotype label rather than a genetic entity: the historical HMSN classification used it for CMT accompanied by pyramidal signs, and families answering that description have turned out to be genetically heterogeneous, with pyramidal involvement arising in more than one dominant disease. No single gene attaches to the designation, so the literature supports no mechanism for it as a unit.

Rationale: Heterogeneity, not an unmapped locus, is what blocks a mechanism here. The designation collects diseases that share a clinical picture and differ in cause, so any allele class named for the group would in fact belong to one of its members, and nothing generalizes to HMSN-5 itself.

Learn more about HMSN-5 →

Mechanistic basis: Gene unknown

Prediction: Chronic cough and gastroesophageal reflux accompanying dominant sensory loss give HSAN-1B a recognizable clinical signature, and linkage in the reported pedigree places the disease on chromosome 3p22-p24. No gene within that interval has been confirmed, and with no transcript and no segregating alleles to examine, the literature supplies no protein on which a mechanism could act.

Rationale: The mapped interval carries many candidate transcripts and no confirmed one, so the evidence fixes a location and goes no further. The cough and reflux phenotype anchors the entity clinically and sets it apart from SPTLC1-related HSAN1A at 9q22, but a phenotype is not a substitute for an allele class.

Learn more about HSAN-1B →

Mechanistic basis: Gene unknown

Prediction: Linkage in a dominantly transmitted sensory neuropathy kindred places HSN-1B in an interval at 3p22-p24, and the clinical picture combines adult-onset sensory loss with chronic cough and gastroesophageal reflux. No gene within the interval has been cloned and no segregating allele has been characterized, so the literature identifies a map position rather than a protein and supports no molecular mechanism.

Rationale: A mechanism requires an allele acting on a product, and HSN-1B has neither on record: the interval holds multiple transcripts and none confirmed to track with the phenotype. Dominant transmission does constrain what the eventual gene could be doing, since one altered copy suffices, but a mode of inheritance is not a mechanism.

Learn more about HSN-1B →