KDM6A + STAG2
KDM6A and STAG2 co-occurrence in bladder cancer (MSK-IMPACT 50K)
DataMSK-IMPACT 50K (2026) · Bladder cancer · 1,917 tumours · both altered in 85 (49 expected if independent)
Strengthened. Significant, with a real effect in a consistent direction, and it still holds when tumours are compared only within the same tumour type (and within the same molecular subtype, wherever this cancer's subtypes are known in the data). No artifact check failed.
A well-studied pair (311 papers): on the map it is known biology, a positive control rather than an overlooked discovery.
Not called novel: a PubTator3 search finds 12 papers naming both KDM6A and STAG2 in bladder cancer in the title or abstract. See the PubTator3 search.
Signal (x) is the corrected significance on the map's scale, discounted from 56% to 25% because the cohort-composition confound is not yet controlled. Efficacy (y) is credibility times actionability.
These genes are altered together more often than chance would predict.
Bladder
This finding's tumour type arises in the highlighted organ.
No structure profile is curated for this gene.
A human gene in this finding. Not in the curated driver set, so no structured profile is shown.
KDM6A mutations in bladder cancer →KDM6A co-mutations by tumour type →Other findings with KDM6A →No structure profile is curated for this gene.
A human gene in this finding. Not in the curated driver set, so no structured profile is shown.
STAG2 mutations in bladder cancer →STAG2 co-mutations by tumour type →Other findings with STAG2 →Anatomical figure from the EBI Expression Atlas anatomogram, and predicted structures from the AlphaFold Protein Structure Database (both CC-BY-4.0). A model shows the protein's healthy fold; a highlighted residue marks a recurrent cancer position, not a prediction of a specific mutation's effect.
The raw 2×2 the FDR q and log₂ odds ratio above were computed from. It is the immutable evidence the score re-derives from: every number on this page traces back to these four counts.
| Gene state | STAG2 altered | STAG2 wild-type |
|---|---|---|
| KDM6A altered | 85 | 416 |
| KDM6A wild-type | 101 | 1,315 |
both 85 (expected 49 if independent) · KDM6A-only 416 · STAG2-only 101 · neither 1,315
Strong evidence. Exceeds the α=0.05 reject threshold (e≥20). Unlike a single-look p, this stays valid under the engine's continuous re-testing.
Molecular-subtype certification is unresolved. Recorded lineage only. Within-lineage molecular subtype remains unresolved. The association remains visible, but does not qualify as a subtype-controlled discovery.
The two axes are variables: signal (x) and efficacy (y = credibility × actionability). A point high on both lands in the frontier (top-right). A strong signal with low credibility sits low and to the right: statistically strong, not worth pursuing.
A language model wrote this from the evidence listed below. It is annotation: the engine's verdict above comes only from the measured statistics, never from this text.
The most likely explanation is residual bladder-cancer molecular-subtype or lineage composition: the KDM6A–STAG2 co-occurrence may be enriched in a subtype not resolved by the available lineage adjustment. Ranked alternatives are: (1) residual molecular-subtype composition, (2) cohort-specific artifact or winner's curse, (3) a shared upstream driver, (4) genuine within-tumour cooperation, and (5) deterministic co-amplification of neighbouring genes. This is a population-level hypothesis, not a claim that the alterations cooperate within individual tumours.
- 1The population-level association is substantial enough to investigate: it survives multiple-testing correction, is burden-aware, has a positive lineage-stratified effect, and includes 85 double-altered tumours. These statistics establish neither a mechanism nor within-tumour interaction.E1E2E18
- 2Residual molecular-subtype composition is the leading explanation because lineage is recorded but molecular subtype remains unresolved. The corpus lead CONVERGES most directly with its prior KDM6A–STAG2 bladder interpretation and is directionally consistent with its subtype-composition interpretation of another bladder co-occurrence; the glioma and lung findings are contextual STAG2 precedents, not direct evidence for this pair.E2E14E15E16E17
- 3A cohort-specific artifact or winner's curse is the strongest competing explanation: the association has not recurred across independent patient sources. A shared upstream driver is possible but has no direct support here; deterministic co-amplification is also unsupported by the supplied evidence. The adjusted association does not rule out subtype confounding.E1E2E3
- 4The supplied paper abstracts do not establish a study of the specific KDM6A–STAG2 relationship: the bladder paper describes a broad comparative genomic cohort, while the other abstract concerns papillary renal cell carcinoma. The co-mention count is therefore not evidence of a mechanism; the lack of a supplied pair-specific study strengthens the claim that a within-tumour interpretation remains under-explored.E8E9E10
Test the leading explanation in an independent bladder-cancer cohort with molecular-subtype annotation: estimate the KDM6A–STAG2 association within subtypes and in a model adjusted for subtype, lineage, alteration burden, and assay coverage. Attenuation of the association after subtype adjustment would support residual subtype composition. To earn a within-tumour claim, use single-cell or clonal DNA sequencing to establish that both alterations occur in the same tumour clone, followed by a functional co-dependency test.
Confidence is limited because replication is currently confined to one patient source and molecular-subtype composition is unresolved. The pattern could be cohort-specific or a winner's-curse artifact despite its corrected significance. No supplied Europe PMC abstract actually studies this gene pair, so a cooperation or redundancy mechanism cannot be responsibly specified; population co-occurrence alone does not establish one.
Weakest link: No subtype-resolved analysis shows that KDM6A–STAG2 co-occurrence is concentrated in a particular subtype, and the signal lacks independent-cohort replication, so residual subtype composition is plausible but not established over cohort artifact or winner's curse.
Significant (after FDR correction) in one independent source of this tumour type so far. The engine records where a pattern is significant, not every source that could test it, so this calls for caution: a pattern selected from many tests tends to overstate its effect. It does not count against the pattern.
- E1In Bladder cancer (MSK-IMPACT 50K 2026), KDM6A and STAG2 are co-occurring: burden-aware DISCOVER p = 4.6e-8 (a test that accounts for each tumor's overall alteration burden, so co-occurrence is not inflated by unstable tumors carrying more of everything), Benjamini-Hochberg FDR q = 5.1e-5, within-stratum (Mantel-Haenszel) log2 odds ratio 1.4 (95% CI 1.0 to 1.8), 85 tumors carry both alterations.cBioPortal / MSK-IMPACT 50K (2026)
- E2The co-occurrence was tested at adjusted FDR q=5.1e-5 with 100% annotation coverage. Recorded lineage only. Within-lineage molecular subtype remains unresolved. The measured heterogeneity test did not reject a common effect; this does not prove homogeneity. This control alone establishes neither a mechanism nor novelty.cBioPortal / MSK-IMPACT 50K (2026), scoped DISCOVER control
- E3This co-occurrence is FDR-significant in only 1 independent patient source so far and does NOT recur across sources, so it could be cohort-specific biology or a winner's-curse artifact of scanning many cohorts. Treat confidence accordingly.cBioPortal (cross-cohort)
- E4KDM6A expression is lower in tumor than matched GTEx+adjacent normal tissue: log2 fold change -0.22 (10.4 vs 12.1 TPM, n=407 tumor / 28 normal).UCSC Xena (tumor vs GTEx normal)
- E5KDM6A strongest disease associations (overall score 0–1): Kabuki syndrome 2 0.80; Kabuki syndrome 0.75; urinary bladder cancer 0.70; urinary bladder carcinoma 0.62.Open Targets
- E6No approved drug directly targets KDM6A (a tractability gap).Open Targets
- E7Tractability for KDM6A: Structure with Ligand, High-Quality Ligand, Med-Quality Pocket, Database Ubiquitination.Open Targets
- E8311 paper(s) co-mention this: well studied. Against each gene's own cancer literature, ~2 co-mentions would be expected by chance, so this pair is already studied more than its parts predict. Most relevant: "Comparative genomic landscape of primary and metastatic bladder urothelial carcinoma in a…" (2026).Europe PMC
- E9"Comparative genomic landscape of primary and metastatic bladder urothelial carcinoma in a large-scale cohort." (2026): Background: Metastatic bladder urothelial carcinoma has poor survival, and large comparative genomic studies using uniform targeted sequencing of paired primary and metastatic lesions remain limited. We compared gene- and pathway-level alterations between primary and metastatic tumors METHODS: We analyzed 2,880 bladder urothelial carcinoma samples (2…Europe PMC (paper)
- E10"Papillary renal cell carcinoma, formerly known as Type 2: a single institutional study addressing histologic and molecu…" (2026): Aims: Papillary renal cell carcinoma (pRCC) accounts for 15%-20% of RCC cases and is the second most common histologic subtype of RCC. In contrast to other common RCC subtypes, there continues to be ongoing debate about how to classify RCCs with papillary architecture and eosinophilic cytoplasm given the heterogeneity of histologic, IHC and molecular…Europe PMC (paper)
- E110 clinical trial(s) match KDM6A/STAG2 in Bladder cancer, 0 currently open.ClinicalTrials.gov
- E12KDM6A: tumor suppressor, driver in 18 cancer types (intOGen 2024.09.20)intOGen (driver catalogue)
- E13STAG2: tumor suppressor, driver in 11 cancer types (intOGen 2024.09.20)intOGen (driver catalogue)
- E14Prior finding (KDM6A + STAG2 in Bladder cancer): verdict inconclusive, confidence 50%. “The most likely explanation is unresolved bladder-cancer molecular-subtype composition: KDM6A and STAG2 alter…”Cybergenic corpus
- E15Prior finding (IDH1 ⊥ STAG2 in Glioma): verdict inconclusive, confidence 49%. “The IDH1–STAG2 mutual exclusivity in this glioma cohort is more likely to reflect molecular-subtype or lineag…”Cybergenic corpus
- E16Prior finding (DOT1L + STAG2 in Non-small cell lung cancer): verdict inconclusive, confidence 20%. “DOT1L + STAG2 in Non-small cell lung cancer: FDR q=3.0e-1 controlled on lineage (pooled q=5.6e-1), log2 odds …”Cybergenic corpus
- E17Prior finding (FGFR3 + PIK3CA in Bladder cancer): verdict inconclusive, confidence 52%. “The FGFR3–PIK3CA co-occurrence is most likely driven by residual molecular-subtype or lineage composition wit…”Cybergenic corpus
- E18Significance is FDR-corrected: Benjamini-Hochberg q = 5.1e-5 across 9870 gene pairs tested in this cohort, so this pair SURVIVES multiple-testing correction (q ≤ 0.05). The q, not the nominal p, is the honest signal strength.Cybergenic method
Model-written annotation, not a finding. The statistics it cites are real and are shown above; the argument connecting them is a hypothesis that needs wet-lab or clinical validation before it means anything.
- Permalink
- https://cybergenic.im/findings/95fd8d66-6f4c-4be7-9268-4d8d7a93b11a
- First seen
- 29 Sep 2026
- Prediction registry
- Locked 29 Sep 2026, row hash
dcb8c249943387bbdd9ad949750ad9e4(not testable out of sample). How the registry works
@misc{cybergenic_95fd8d66,
author = {{Cybergenic}},
title = {{KDM6A and STAG2 co-occurrence in bladder cancer (MSK-IMPACT 50K)}},
year = {2026},
howpublished = {\url{https://cybergenic.im/findings/95fd8d66-6f4c-4be7-9268-4d8d7a93b11a}},
note = {Research engine finding, first seen 2026-09-29; prediction locked 2026-09-29, registry row hash dcb8c249943387bbdd9ad949750ad9e4},
month = sep
}