RareResearch.AI
← Back to atlas

W639G

Category 4 — Stable Fold, Function DisruptedPathogenicTransmembrane · predictedEditorial
TryptophanGlycine at position 639 · TM10 (632-652), helical transmembrane · WFS1 (Wolframin)

Tryptophan → Glycine at position 639 inside wolframin's tenth transmembrane helix (TM10). ClinVar Pathogenic. AlphaMissense 0.481 (BELOW the likely-pathogenic threshold), DynaMut2 ΔΔG -1.03 kcal/mol (destabilising). A variant where the structural cost is clear but the AlphaMissense signal is unexpectedly weak.

Interactive 3D Structure

Wild-type reference
Wild-type W639 — hydrogen bond to I643
Fullscreen ↗
DynaMut2 mutant · W639G
Mutant G639 — hydrogen bond contact to I636 lost
Fullscreen ↗

Bond changes · DynaMut2 interaction analysis

0 lost2 gained8 preserved
Interaction typeWild-type partnerMutant partnerStatus
Hydrogen bondL635L635Preserved
Hydrogen bondI636I636Preserved
Hydrogen bondA642A642Preserved
Hydrogen bondI643I643Preserved
Polar contactL635L635Preserved
Polar contactI636I636Preserved
Polar contactA642A642Preserved
Polar contactI643I643Preserved
CarbonylI636Gained
Van der WaalsL637Gained

Lost / gained / preserved interatomic contacts at the variant residue, from the DynaMut2 (Arpeggio) interaction analysis of the wild-type and energy-minimized mutant structures.

Computational Predictions

DynaMut2 ΔΔG
-1.03kcal/mol
Destabilising — moderate
AlphaMissense
0.481
Amb
AlphaFold pLDDT
87
model confidence
Schema
Cat 4
Category 4 — Stable Fold, Function Disrupted
ΔΔG confidence: Reduced confidence · Transmembrane
  • transmembrane position; predictor benchmarked on soluble proteins

This variant sits in the transmembrane band. The stability value was computed with a predictor benchmarked on water-soluble proteins. Published benchmarks of ddG predictors on membrane proteins report correlation below 0.4. Treat the magnitude as unreliable and the sign as weak evidence only.

Do not rank this value against a variant from a different region on ΔΔG alone.

Clinical Evidence

ClinVar classificationPathogenic
Review statuscriteria provided, single submitter
Associated conditions(no specific conditions catalogued for W639G — ClinVar Pathogenic by review evidence)
InheritanceInheritance not specified.
Population frequency (gnomAD v4)Ultra-rare · AF 0.000068%
cDNA changec.1915T>G
ClinVar accessionVCV003720568
Last evaluated2024/10/03 00:00

Observed at very low frequency in gnomAD.

Classified for1★ unverified submission

ClinVar classifies this variant as Pathogenic, but does not state what it is classified for. ClinVar records the condition only as "not provided". A classification without a phenotype and an inheritance mode cannot be read as a statement about Wolfram syndrome risk or severity.

  • Phenotype scope is interpretation of submitted records. The computed values on this card (AlphaMissense, DynaMut2 ΔΔG, pLDDT) are measurements of the model, not of a patient.
  • Presentation cannot be generalised from neighbouring variants: one nucleotide over can present very differently. Structural similarity supports a mechanistic hypothesis, never a phenotype claim.

Review status: 1★ single submitter. ClinVar "conditions" is the union of all submissions; 0–1★ entries are submitted conditions, not documented phenotypes.

Population frequency
Global (all ancestries)
AF 0.000068% · 1 / 1,461,726 alleles
Homozygotes
0

Every ancestry group with observed alleles falls below the reporting threshold (min 2 alleles, min 500 sampled) — too sparse to name a highest-frequency population.

No homozygotes reported in gnomAD v4. For a recessive allele this is expected and is not, on its own, evidence either way.

Ancestry groupAllele freq.AC / ANHom.Carrier est.
European (non-Finnish) · under-sampled0.000090%1 / 1,112,0100

Source: gnomAD r4, retrieved 2026-08-21. Values are allele frequencies. The carrier estimate is derived from the allele frequency (Hardy-Weinberg, ~2·AF) and is an approximation, not a published figure — where a published carrier frequency exists it outranks this column and is cited on the card.

Structural Context

Position 639 sits in the middle of TM10. The AlphaFold model places W639 within 5 Å of VAL638 (2.5 Å), LEU640 (2.5 Å), ILE636 (3.6 Å), LEU635 (3.8 Å), and ALA642 (4.3 Å). The local environment is uniformly hydrophobic — an aliphatic-rich pocket where the wild-type tryptophan's bulky indole ring contributes substantial volume.

Replacing tryptophan with glycine removes the indole ring entirely, leaving a small cavity in the TM10 hydrophobic core. The DynaMut2 |ΔΔG| of 1.03 captures this — the volume mismatch produces meaningful destabilization, larger than most variants in this batch.

Yet AlphaMissense places this at 0.481 — below the 0.564 likely-pathogenic threshold. AM considers this variant likely benign. The discrepancy with ClinVar Pathogenic classification is striking.

Two interpretations: (1) AM's training data may under-represent TM-helix variants in general, leading to systematic under-calling of pathogenicity in this structural context; (2) The variant may be technically pathogenic in specific clinical scenarios (compound heterozygosity, specific tissue contexts) but not in the more general population-genetics sense AM is trained on. The DynaMut2 destabilization (|ΔΔG| 1.03) is real and substantial.

This is a variant where ΔΔG is informative and AM is uncertain.

Amino-acid chemistry
Tryptophan (W) → Glycine (G) — bulky aromatic replaced by the smallest amino acid. Massive volume loss; aromatic character entirely eliminated.
Position in the protein
TM10 (residues 632–652) · position 639 is in the middle of TM10, bilayer-embedded (pLDDT 86).

Druggability Assessment

Category 3/4 — Most Druggable (AM caveat). |ΔΔG| = 1.03 kcal/mol — fold absorbs the substitution at meaningful cost. AlphaMissense 0.481 is below the likely-pathogenic threshold but ClinVar Pathogenic classification is established.

The mechanism is volume loss in a TM10 hydrophobic packing pocket — the lost tryptophan indole creates a cavity that destabilizes the local membrane-embedded geometry. Therapeutic strategy: site-directed small molecules that fill the cavity left by W639, restoring the wild-type packing volume.

The unexpected AM signal mismatch is itself a finding the Atlas surfaces — variants where structural destabilization is clear but AM training under-calls pathogenicity exist, and they deserve wet-lab characterization to clarify mechanism.

Why this matters

W639G demonstrates the value of the Atlas's dual-metric framing in the opposite direction from variants like T361I. Where T361I has stabilising ΔΔG but high AM (functional mechanism), W639G has destabilising ΔΔG but low AM (structural mechanism not captured by AM training). The Atlas captures both — the framework is more complete than either metric alone.
Therapeutic Strategy Handoff · prediction

Feed this card to Wolfram Intelligence

Download the W639G PDF below and upload it to Wolfram Intelligence to generate therapeutic-strategy proposals — guanidinium mimetics, sigma-1 agonist docking, NAC thiol-capping. NAC is already on the bench-testing list.

Download W639G PDF card ↓Strategies are AI-generated predictions, not validated therapeutics.

UniProt Domain Annotation

Chain1890 · Wolframin
Transmembrane632652 · Helical