Oncofusion

start / graph / fusion-yap1-kmt2a-yap1

Fusionstrong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.Q92-kmt2a-sarcoma

YAP1::KMT2A::YAP1 (YKY) - an insertional KMT2A rearrangement that keeps the CXXC zinc finger and throws the menin-binding exon away

also called YKY · YAP1-KMT2A-YAP1 · YAP1::KMT2A · KMT2A::YAP1 · YAP1-KMT2A

In a rare soft-tissue cancer of young adults, a small piece of the KMT2A gene is cut out and pasted into the middle of the YAP1 gene, instead of the usual head-to-tail gene fusion. The piece that moves carries only KMT2A's DNA-gripping CXXC finger, and leaves behind the very first exon - the one that holds the grip site for menin, the protein that menin-inhibitor drugs act on.

The same thing, for a scientist

YAP1::KMT2A::YAP1 is the commonest KMT2A rearrangement in soft-tissue sarcoma and is architecturally unlike a leukaemic KMT2A fusion in three ways at once. (i) It is an INSERTION, not a 5'-3' fusion: KMT2A exons 4/5-6 are inserted between YAP1 exons 4/5 and 8/9, replacing YAP1 exons 5/6-8, verified by RT-PCR in three SEF-like sarcomas [E1]. (ii) KMT2A is the 3' or internal contributor, not the 5' partner, and its contribution is only the CXXC-type zinc finger (aa 1147-1195, exons 5-6) - the menin-binding motif (aa 6-25) and both LEDGF/PSIP1 integrase-binding motifs sit in KMT2A exons 1-2 and are absent from the insert [E2][E4]. (iii) The repression domain 2 required for PAF1 complex engagement (exons 7-8) is also absent, and the tumours show no HOXA9/MEIS upregulation [E2][E5]. YAP1 contributes its TEAD-binding N-terminus and its C-terminal PDZ-binding motif, and the transcriptional output of YKY constructs in immortalised fibroblasts overlaps other YAP1 fusions with Wnt and Hedgehog pathway enrichment [E1][E2]. The honest caveat is that reciprocal KMT2A::YAP1 transcripts with 5' KMT2A breakpoints at exons 3, 4 and 6 - which DO retain exon 1 - are detected in a substantial minority of cases [E3], so this is a strong but not clean natural experiment on the menin interface.

Go deeper
The architecture, established step by step

Step 1 - it is not a simple fusion. Two reciprocally oriented breakpoints between YAP1 and KMT2A were found in every fully evaluated case, together with concurrent deletion of the same KMT2A exons implicated in the fusion. YAP1 (11q22.1-22.2) lies about 16 Mb centromeric to KMT2A (11q23.3) on the same strand and in the same transcriptional direction, so a reciprocal translocation cannot explain the two breakpoint sets, and a simple interstitial deletion cannot either. The authors infer a complex insertional (cut-and-paste) allele 2.

Step 2 - the insert is named. RT-PCR across three SEF-like sarcomas put KMT2A exons 4/5-6, encoding the CXXC domain, between YAP1 exon 4/5 and exon 8/9, replacing YAP1 exons 5/6-8 1.

Step 3 - the largest series agrees on the exon range. Across 16 YAP1::KMT2A fusions with known breakpoints, the breaks involve YAP1 exons 4-6 and KMT2A exons 4-6, the commonest being YAP1(ex5)::KMT2A(ex4) in five cases and YAP1(ex4)::KMT2A(ex5) in five 3.

Step 4 - translate the exons into residues. KMT2A exons 5-6 encode aa 1112-1212, which contains the CXXC-type zinc finger (aa 1147-1195) and nothing else of note. The menin-binding motif (aa 6-25) and IBM1 (aa 123-134) are in exon 1; IBM2 (aa 147-152) is in exon 2; all three AT-hooks are in exon 3 4.

Two variants, counted - and one breakpoint that lands inside the CXXC

Massoth's Results resolve the architecture per case, using YAP1 NM_001130145 and KMT2A NM_001197104 6:

variant Ivariant II
cases13 of 15 fully evaluated2 of 15
junction 1YAP1 ex5 :: KMT2A ex4YAP1 ex3 :: KMT2A ex5
junction 2KMT2A ex6 :: YAP1 ex9KMT2A ex6 :: YAP1 ex5
readinginsert of KMT2A exons 4-6 into YAP1, losing YAP1 exons 6-8insert of KMT2A exons 5-6 between YAP1 exons 3 and 5
separate KMT2A exon deletionexons 4-6 in 8 of 13exons 5-6 in 1 of 2

Two things follow that matter more than the counts.

First, the 'reciprocal' transcript is mostly not reciprocal. Junction 2 is the downstream junction of the insert, so a transcript reported as KMT2A(ex6)::YAP1(ex9) in these cases is part of the YKY allele and begins at YAP1 exon 1, not at KMT2A exon 1. This project's first pass treated every reported KMT2A::YAP1 junction as a possible menin-competent 5'-KMT2A species; for Massoth's 15 cases that is wrong, and the menin-motif-absent conclusion is correspondingly stronger than first written. It remains open only for the KMT2A(ex3)::YAP1(ex5) and KMT2A(ex4)::YAP1(ex4) breakpoints reported in a later series, which match neither insert junction 3.

Second, the CXXC is not always intact. Massoth's Table 1 footnote records a breakpoint inside KMT2A exon 5 at amino acid 1143, within what the authors call the 'pre-CxxC' region (aa 1115-1154) that "accounts for a portion of the multivalent bond with PAFc" 6. Amino acid 1143 is four residues short of the CXXC zinc finger's start at 1147, so that allele splits the pre-CxxC/PAF1c element and truncates the CXXC. This is the case Kao's phrase 'a partially disrupted or totally lost DNA binding domain' is describing, and it is why that phrase and Massoth's 'preservation of KMT2A CxxC-binding domain' are not in conflict - see the reconciliation in concept-kmt2a-fusion-polarity-is-inverted-in-sarcoma.

What the fusion protein does and does not contain
elementresiduesin YKY?
KMT2A menin-binding motif6-25 (exon 1)NO
KMT2A IBM1 / IBM2 (LEDGF)123-134 / 147-152 (exons 1-2)NO
KMT2A AT-hooks 1-3169-309 (exon 3)NO
KMT2A CXXC zinc finger1147-1195 (exons 5-6)YES
KMT2A repression domain 2 (PAF1c)exons 7-8NO
KMT2A PHD / bromodomain / FYR / SETexons 11-36NO
YAP1 TEAD-binding domainexon 1YES
YAP1 PDZ-binding motifexon 9YES
YAP1 exons 5/6-8 (part of the transactivation region)replaced by the KMT2A insert

Assignments: KMT2A side from 4; YAP1 side as stated by the primary reports 12.

What this implies for the index fusion, by contrast

This project's model of KMT2A::AFF1 is: retain the KMT2A N-terminus, keep menin and LEDGF bound at the MBM/IBM interface, keep the CXXC domain addressing unmethylated CpG, and replace the C-terminus with an elongation-competent partner (AFF1 -> SEC -> P-TEFb -> DOT1L). YKY keeps exactly one of those four elements. It has the CpG address and nothing else: no menin interface, no LEDGF interface, no AT-hooks, no RD2 and therefore no PAF1c engagement, and a partner that is a Hippo transcriptional co-activator rather than an elongation factor. The tumours do not upregulate HOXA9 or MEIS 5. If the menin interface were merely one of several redundant routes into the leukaemic program, one would expect a HOXA-driven phenotype here; the absence of one is consistent with the project's position that the menin-bound N-terminus is the determinant - and it also means this entity is the wrong place to test a menin inhibitor.

Where the natural experiment is not clean

Reciprocal KMT2A::YAP1 transcripts are reported alongside the YAP1::KMT2A transcript in most series. In the largest series, six had known breakpoints and the 5' KMT2A end was exon 3, 4 or 6 - all of which include exon 1 and therefore the menin-binding motif 3. One series detected the reciprocal transcript in all three index cases by RNA sequencing and confirmed it by RT-PCR in one 5. Under the insertional model these reciprocal transcripts are by-products of the same rearrangement rather than independent drivers, and their breakpoints are discrepant with the forward transcript's - which is the argument for the single YKY allele in the first place 2. But no experiment has shown that the reciprocal product is not expressed as protein, and none has tested whether it binds menin. Until that is done, the correct statement is that the dominant species lacks the menin interface, not that the tumour contains no menin-competent KMT2A species.

Facts and figures
gene5YAP1
gene3KMT2A
architectureinsertional: YAP1 exons 1-4/5 + KMT2A exons 4/5-6 + YAP1 exons 8/9-9 (a three-segment allele, not a 5'-3' fusion)
cytobandintrachromosomal 11q22.1-22.2 / 11q23.3; the two genes lie ~16 Mb apart on the same strand and in the same transcriptional direction
uniprot5P46937
uniprot3Q03164
length AA 5504
length AA 33969
menin binding motif in fusionfalse
ledgf ibm in fusionfalse
cxxc in fusionretained in the common configurations (KMT2A exons 4-6 or 5-6), but NOT universally: one reported breakpoint falls inside exon 5 at aa 1143, within the pre-CxxC PAF1c element and 4 residues before the CXXC zinc finger's start at aa 1147
set domain in fusionfalse
detectionFISH is insensitive - break-apart probes for KMT2A and/or YAP1 were negative in cases confirmed by sequencing; RNA-based NGS is the reliable assay
transcripts used by the sourceYAP1 NM_001130145 and KMT2A NM_001197104, stated in the paper's Methods - the same 36-exon KMT2A frame this graph's exon map uses
stage at presentation13 of 16 stage IV (81%), 1 stage III, 2 stage II
sex yap1 arm56% female in the 16-case YAP1 arm, which disagrees in direction with the 1.4:1 male ratio of the 22-case series
diseases
[
  "KMT2A-rearranged soft tissue sarcoma",
  "MUC4-negative sclerosing epithelioid fibrosarcoma"
]
breakpoints
[
  {
    "id": "bp-yap1-ex5-kmt2a-ex4",
    "gene5_exon_last": 5,
    "gene3_exon_first": 4,
    "frequency_pct": 31,
    "population": "16 YAP1::KMT2A fusions with known breakpoints, Warmke 2025 series",
    "domains_retained_5": [
      "YAP1 N-terminus (exon 1), which carries the TEAD-binding region - no UniProt domain feature exists for it, so it is named positionally rather than as a domain with asserted coordinates"
    ],
    "domains_lost_5": [
      "part of the YAP1 transactivation region (exons 5/6-8, replaced by the KMT2A insert)"
    ],
    "domains_retained_3": [
      "CXXC"
    ],
    "domains_lost_3": [
      "MBM (menin-binding motif)",
      "IBM1 (integrase-binding motif 1)",
      "IBM2 (integrase-binding motif 2)",
      "AT-hook 1",
      "AT-hook 2",
      "AT-hook 3",
      "repression domain 2 (RD2, KMT2A exons 7-8)",
      "PHD1",
      "PHD2",
      "PHD3",
      "bromodomain",
      "FYRN",
      "FYRC",
      "SET"
    ],
    "phenotype_notes": "no HOXA9 or MEIS upregulation reported in the YAP1 arm; Wnt and Hedgehog pathway genes enriched in YKY-expressing tumours and fibroblasts"
  },
  {
    "id": "bp-yap1-ex4-kmt2a-ex5",
    "gene5_exon_last": 4,
    "gene3_exon_first": 5,
    "frequency_pct": 31,
    "population": "16 YAP1::KMT2A fusions with known breakpoints, Warmke 2025 series",
    "domains_retained_3": [
      "CXXC"
    ],
    "domains_lost_3": [
      "MBM (menin-binding motif)",
      "IBM1",
      "IBM2",
      "AT-hooks 1-3",
      "repression domain 2 (RD2, KMT2A exons 7-8)",
      "SET"
    ],
    "phenotype_notes": "second most common configuration, tied with YAP1(ex5)::KMT2A(ex4) at 5 of 16"
  }
]
reciprocal transcripts
{
  "detected": true,
  "breakpoints_reported": [
    "KMT2A(ex6)::YAP1(ex9) x2",
    "KMT2A(ex6)::YAP1(ex7) x2",
    "KMT2A(ex3)::YAP1(ex5) x1",
    "KMT2A(ex4)::YAP1(ex4) x1"
  ],
  "menin_motif_status": "RETAINED in the reciprocal products, because their 5' KMT2A end begins at exon 1 regardless of where it stops",
  "note": "REVISED after reading Massoth's Results rather than its abstract. In the 15 fully evaluated cases the two breakpoint sets are the TWO JUNCTIONS OF ONE INSERTIONAL ALLELE (YAP1ex5|KMT2Aex4 and KMT2Aex6|YAP1ex9), so a 'KMT2A::YAP1' junction there does NOT imply a transcript beginning at KMT2A exon 1 and does NOT restore the menin motif. What remains genuinely open is the subset of breakpoints reported elsewhere as KMT2A(ex3)::YAP1(ex5) and KMT2A(ex4)::YAP1(ex4), which do not match either insert junction and would be 5'-KMT2A products containing exon 1."
}
frequency
{
  "share_of_kmt2a_rearranged_sarcoma_pct": 47,
  "n_of_cohort": "16 of 34 (Massoth 2020)"
}
variants
{
  "variant I": {
    "cases": "13 of 15 fully evaluated (patients #2-14)",
    "junction_1": "YAP1 exon 5 :: KMT2A exon 4",
    "junction_2": "KMT2A exon 6 :: YAP1 exon 9",
    "interpretation": "insertion of KMT2A exons 4-6 into YAP1 with loss of YAP1 exons 6-8",
    "separate_kmt2a_deletion": "exons 4-6, detected at DNA and/or RNA level in 8 of the 13"
  },
  "variant II": {
    "cases": "2 of 15 (patients #15-16)",
    "junction_1": "YAP1 exon 3 :: KMT2A exon 5",
    "junction_2": "KMT2A exon 6 :: YAP1 exon 5",
    "interpretation": "insertion of KMT2A exons 5-6 into YAP1 between exons 3 and 5",
    "separate_kmt2a_deletion": "exons 5-6 in one of the two"
  }
}
pre cxxc element
{
  "residues": [
    1115,
    1154
  ],
  "what_it_is": "the 'pre-CxxC' region, which the source names as accounting for a portion of the multivalent bond with the PAF1 complex",
  "why_it_matters": "one reported breakpoint falls INSIDE KMT2A exon 5 at amino acid 1143, i.e. within this element and 4 residues short of the CXXC zinc finger - so in that case the pre-CxxC/PAF1c element is split and the CXXC itself is truncated",
  "source": "Massoth et al. 2020, Table 1 footnote (PMID 32461620)"
}
co alterations yap1 arm
{
  "homozygous CDKN2A loss": 8,
  "homozygous RB1 loss": 2,
  "MDM2 amplification": 2,
  "of_n_tumours": 16,
  "median_TMB_mut_per_Mb": 1.2,
  "TMB_range": "<0.8-5.7",
  "note": "recorded from Massoth's Results; this is the YAP1 arm's co-alteration profile and it is CDKN2A-dominated, unlike the VIM arm's CTNNB1/SMARCB1/ARID1A pattern"
}
How this could bite us back
  • Menin inhibitors are the only KMT2A-directed drugs that exist, and by this architecture they have no binding site in the dominant fusion species. A KMT2A-rearranged sarcoma referred for a menin inhibitor on the strength of the gene name alone would be a predictable failure.
  • If the reciprocal KMT2A::YAP1 product IS menin-competent, a menin inhibitor would hit a species that is not the driver - activity in a model would then be uninterpretable and absence of activity in a patient would not exonerate the target.
  • The functional output points at YAP1/TEAD, so the plausible drug is a TEAD palmitoylation inhibitor rather than anything KMT2A-directed; but YAP1's C-terminal transactivation region is partly replaced in this allele, and no TEAD inhibitor has been tested in this entity.
  • DOT1L inhibition, this project's other KMT2A-directed modality, has an even weaker rationale here: the fusion carries no elongation-machinery partner and the tumours show no HOXA output for DOT1L to sustain.
  • The YAP1 arm carries homozygous CDKN2A loss in eight of sixteen tumours and RB1 loss or MDM2 amplification in two each. Any fusion-directed single-agent proposal in this arm is competing with a cell-cycle lesion in half the cases, and a CDK4/6 rationale has at least as much genomic support as anything KMT2A-directed.
Still open
  1. Has a YKY protein ever been observed? Every claim about this fusion rests on transcript structure; no western blot, immunoprecipitation or mass-spectrometric identification of the chimeric protein appears in any paper read for this beat.
  2. Does the YKY protein bind menin? The prediction from the exon map is no. A pulldown of menin from a YKY-expressing fibroblast line, against a KMT2A::AFF1-expressing control, is a one-week experiment that would settle it.
  3. Is the reciprocal KMT2A::YAP1 product translated, and does it bind menin? If it is, the entity contains a menin-competent KMT2A species and the natural experiment is void.
  4. Does the retained CXXC zinc finger do anything in this fusion, or is it a passenger carried by whatever structural mechanism cut and pasted the locus? The authors argue for a key role from its consistent retention; consistent retention is also what a fixed breakpoint region produces.
  5. Does the aa-1143 allele behave differently from the intact-CXXC alleles? It splits the pre-CxxC PAF1c-binding element and truncates the zinc finger, so it is the internal negative control for the claim that CXXC retention matters here - and its case is buried in a table footnote.
Every source behind this entry (6)
  1. E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    In three human soft-tissue sarcomas with sclerosing-epithelioid-fibrosarcoma-like morphology, RT-PCR verified that a portion of KMT2A (exons 4/5-6) encoding the CXXC domain is inserted between YAP1 exon 4/5 and exon 8/9, replacing YAP1 exons 5/6-8, and expression profiling of YKY-expressing tumours and immortalised fibroblasts showed enrichment of Wnt and Hedgehog pathway genes overlapping other YAP1 fusions.

    PMID 37246732DOI
    What this evidence cannot show

    Human tumour tissue plus an immortalised-fibroblast overexpression model, not a sarcoma-lineage or haematopoietic system, and n=3 tumours for the RT-PCR. Full text is not readable: Europe PMC returns no PMCID, the fullTextXML, efetch db=pmc and PMC HTML routes are therefore unavailable, and the Wiley DOI returns HTTP 403. Everything recorded here comes from the abstract, so the exact exon usage per case and the fibroblast transformation readouts were not read.

    Evidence that argues against this (1)
    • The same study also built simple YAP1::KMT2A and KMT2A::YAP1 constructs and studied their effects, so the authors did not treat the two-part fusions as biologically irrelevant. Whether the insertional YKY is the sole pathogenic species or the most common of several is not settled by this paper's abstract. Massoth LR et al. 2020
  2. E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    In a comprehensive genomic profiling database of 14,680 human sarcomas, YAP1-KMT2A rearrangements showed two reciprocally oriented breakpoints plus concurrent deletion of the same KMT2A exons in every fully evaluated case, consistent with a complex YAP1-KMT2A-YAP1 allele that retains the CxxC-binding domain in KMT2A exons 5-6 but lacks the repression domain 2 encoded by exons 7-8; the authors note that no increase in HOXA-related gene expression had been detected in three previously reported YAP1-KMT2A sarcomas.

    PMID 32461620DOI
    What this evidence cannot show

    The YKY configuration is an inference from DNA and RNA breakpoint pairs and from concurrent exon-level copy loss, not a sequenced single allele or a cloned cDNA spanning the whole insert; the authors state it as a hypothesis. The cohort is a commercial profiling database enriched for advanced, aggressive disease, so the case mix is not population-based. The HOXA statement is the authors' reading of another group's RNA-seq, not their own measurement.

    Evidence that argues against this (1)
    • Prior studies on the same entity assumed two separate reciprocal transcripts and hypothesised the YAP1-KMT2A fusion alone to be pathogenic, and one of them detected the reciprocal KMT2A-YAP1 transcript in all three of its index cases. The insertional model displaces that reading but has not disproved it. Kao YC et al. 2020
  3. E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    In the largest series of human YAP1::KMT2A-rearranged sarcomas to date (22 cases), the 16 YAP1::KMT2A fusions with known breakpoints all involved YAP1 exons 4-6 and KMT2A exons 4-6, while the six reciprocal KMT2A::YAP1 fusions with known breakpoints involved KMT2A exons 3-4 and 6 joined to YAP1 exons 4-5, 7 and 9; DNA methylation profiling placed the tumours in a cluster distinct from conventional SEF and from low-grade fibromyxoid sarcoma.

    PMID 39641785DOI
    What this evidence cannot show

    Retrospective multi-institutional case collection with referral bias toward diagnostically difficult tumours; breakpoints come from several different NGS assays across institutions rather than one uniform pipeline. The 22-case methylation dataset is stated to be available from the corresponding author on request and was not deposited, so the clustering cannot be independently re-run; the comparison cohorts (GSE110697, GSE112308, GSE140686, EGAS00001007042) are public.

    Evidence that argues against this (1)
    • The reciprocal KMT2A::YAP1 breakpoints reported here at KMT2A exons 3 and 4 place the 5' KMT2A end downstream of exon 1, meaning those alleles DO retain the menin-binding motif. This is the strongest published argument against reading the entity as uniformly menin-interface-free, and it comes from the same series. Kao YC et al. 2020
  4. E4established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database

    Computed here (2026-09-21) by mapping Ensembl transcript ENST00000534358 exon boundaries onto the UniProt Q03164 feature table: KMT2A exons 5-6 encode aa 1112-1212 and therefore contain the CXXC-type zinc finger (1147-1195), while the menin-binding motif (6-25) and IBM1 (123-134) lie in exon 1 and IBM2 (147-152) in exon 2, none of which is contributed to YKY.

    What this evidence cannot show

    A coordinate derivation, not a measurement: it establishes what the reported transcript CAN encode, not what protein the tumour makes. No antibody, mass-spectrometry or western evidence for a YKY protein product exists in any of the papers read for this beat, so the protein has never been observed - only its transcript. The full derivation and the 3-residue Ensembl/UniProt offset check are in ana-kmt2a-exon-to-domain-map.

    Evidence that argues against this (1)
    • The breakpoints are reported as exon indices and some cases break WITHIN an exon rather than at a boundary; one published case carried an out-of-frame 5' KMT2A transcript lacking exon 6, and a low-level splice variant absent of exon 6 was seen in another. An exon-level map cannot resolve what those alleles encode. Massoth LR et al. 2020
  5. E5moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

    In three human MUC4-negative sclerosing epithelioid fibrosarcomas with YAP1-KMT2A fusions, RNA sequencing showed no upregulation of YAP1 or KMT2A (KMT2A was slightly lower than control) and no upregulation of the KMT2A leukaemia target genes HOXA9 and MEIS; the authors note that KMT2A is presumably the 3' partner in these tumours, with breakpoints in exons 4-5 that leave the DNA-binding domain partially disrupted or lost and the SET domain intact.

    PMID 31592798DOI
    What this evidence cannot show

    The HOXA9/MEIS result is reported in the discussion as 'data not shown' with n=3, no housekeeping detail and no deposited data - elink from this PMID to GEO returns zero datasets (checked 2026-09-21), so it cannot be re-analysed. A negative on two genes in three bulk tumour samples is a weak instrument for a transcriptional program; single-cell or ChIP data would be needed to exclude HOXA engagement in a subpopulation.

    Evidence that argues against this (1)
    • The same authors state that in the previously reported VIM-KMT2A sarcoma, upregulation of KMT2A AND of HOXA genes WAS noted. So the HOXA-negative result is specific to the YAP1 arm, and the two morphological subtypes of this entity may differ on exactly the axis that matters for the leukaemia analogy. Yoshida A et al. 2019
  6. E6strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    In Massoth's 16 human YAP1-KMT2A sarcomas, 13 of the 15 fully evaluable cases showed YAP1 exon 5 to KMT2A exon 4 plus KMT2A exon 6 to YAP1 exon 9 (variant I, an insert of KMT2A exons 4-6 with loss of YAP1 exons 6-8) and 2 showed YAP1 exon 3 to KMT2A exon 5 plus KMT2A exon 6 to YAP1 exon 5 (variant II); one breakpoint fell inside KMT2A exon 5 at amino acid 1143, within a 'pre-CxxC' region (aa 1115-1154) the authors state accounts for part of the multivalent bond with the PAF1 complex; concurrent alterations were homozygous CDKN2A loss in eight tumours, homozygous RB1 loss in two and MDM2 amplification in two, with median tumour mutational burden 1.2 mutations/Mb, and 13 of 16 patients presented at stage IV.

    PMID 32461620DOI
    What this evidence cannot show

    Read from the open-access full-text Results and Table 1 footnotes, not the abstract - this beat's first pass used the abstract and understated the architecture. Breakpoints in cases #1-16 are reported by highest read count from targeted RNA and DNA assays, not by cloning the allele, and both variants remain inferences from breakpoint pairs plus concurrent exon-level copy loss. The 81% stage IV figure is a property of a commercial profiling cohort enriched for advanced disease and must not be read as the entity's stage distribution. Six of 16 had orthogonal DNA confirmation; the other ten rest on RNA alone.

    Evidence that argues against this (1)
    • The pre-CxxC element at aa 1115-1154 is named by these authors as contributing to the PAF1c bond, and the amino-acid-1143 breakpoint therefore disrupts it - which cuts against this entry's framing that the YKY allele retains an intact, functional CpG-addressing module. At least one reported allele carries a broken CXXC region, and no functional work distinguishes the intact from the broken form. Kao YC et al. 2020
Connected to