Oncofusion

start / graph / gene-tmprss13

Genemoderate confidence 3 of 5: One strong study, or several smaller consistent ones.Q96-sarcoma-mbm-partners

TMPRSS13 (MSPL) - a cell-surface serine protease fused in frame to KMT2A exon 20, the clearest NOT-APPLICABLE epigenetics call of the seven, and a symbol whose approved form has a third of its alias's recall

also called TMPRSS13 · MSPL · MSP · membrane-type mosaic serine protease · mosaic serine protease · transmembrane protease serine 13 · Q9BYE2

TMPRSS13 is a protease anchored in the cell's outer membrane that snips other proteins outside the cell to switch them on, including the growth factor HGF. In one patient's high-grade endometrial stromal sarcoma KMT2A broke at exon 20 and joined almost the whole of TMPRSS13, which would put a membrane-anchored protease behind a protein whose job is to find DNA.

The same thing, for a scientist

TMPRSS13 (Q9BYE2, 586 aa) is a **type II transmembrane serine protease**: a single-pass signal-anchor at 166-186, a non-catalytic stem with an SRCR domain (195-325) and an LDL-receptor class A module (204-226), and a C-terminal **peptidase S1** domain (326-559) with the classical charge-relay triad at 366, 414 and 511 [E1]. It is a genuine active protease that autocatalytically cleaves its own zymogen, and its substrates are **extracellular proproteins**: it converts pro-**PRSS8/prostasin** and pro-**HGF** to their active forms, the latter promoting MAPK signalling [E2, E3]. Its trafficking is unusually well mapped - N-glycosylation at Asn405 and Asn445 is required to leave the endoplasmic reticulum, phosphorylation is required for surface residency, and the HAI-1/HAI-2 inhibitors (SPINT1, SPINT2) promote that phosphorylation and surface localisation [E2, E3, E4]. **In case S17 the fusion is `KMT2A exon 20 :: TMPRSS13 exon 2`** [E5]. KMT2A supplies aa 1-1888 - MBM, IBM1, IBM2, all three AT-hooks and CXXC - the break being downstream of the leukaemic breakpoint cluster region [E6]. TMPRSS13 exon 2 begins 21 nt after the ATG at a codon boundary, so the junction is predicted **in frame** and the chimera would carry TMPRSS13 aa 8 onward: **the transmembrane signal-anchor, the whole stem and the whole protease domain** [E7]. That is the most complete partner contribution of the seven after DOCK10 - and it is **topologically incoherent**, putting a membrane-inserted, surface-shed protease behind a nuclear CpG-island reader. **Epigenetics: NOT APPLICABLE, and this is the one case in the beat where that is the whole answer.** Four bounded queries return zero or noise and the architecture explains why: a signal-anchored, surface and secreted S1 peptidase whose substrates are extracellular proproteins has no route to a histone [E8]. **Two identifier traps, both measured.** The approved symbol is the LOW-recall one - `TMPRSS13[tiab]` = **38** against `MSPL[tiab]` = **113** - and UniProt lists **TMPRSS11** as a gene synonym, which is not a standalone HGNC symbol at all [E9].

What it looks like
TMPRSS13586 aa
TranTransmembrane signal-anchor (type II) · residues 166–186SRCR (scavenger receptor cysteine-rich)SRCR (scavenger receptor cysteine-rich) · residues 195–325LDL-LDL-receptor class A · residues 204–226Peptidase S1 (trypsin-like serine protease)Peptidase S1 (trypsin-like serine protease) · residues 326–559Active site (charge relay, His366) · residues 366–366Active site (charge relay, Asp414) · residues 414–414Active site (charge relay, Ser511) · residues 511–5110200400

No three-dimensional model is available for this protein, so the map above is the whole picture. That is itself informative: AlphaFold does not model chains this long, and the parts that have been crystallised are held under the entries for the individual domains.

Go deeper
What TMPRSS13 normally does

TMPRSS13 belongs to the type II transmembrane serine protease (TTSP) family. Its topology is the family's: a short cytoplasmic N-terminus, a signal-anchor transmembrane helix at 166-186, a non-catalytic stem carrying an SRCR domain and an LDL-receptor class A module, and a C-terminal serine protease domain facing outward 1. Note the N-terminal 9-93 region is a low-complexity repeat array - thirteen five-residue A-S-P-A-[GLQR] repeats and four T-P-P-G-R repeats - not a functional module.

It is an active protease with two named physiological substrates, both proproteins cleaved outside the cell: pro-PRSS8/prostasin, the epithelial channel-activating protease 3, and pro-HGF, whose activation promotes MAPK signalling 2. In mouse, TMPRSS13 deficiency impairs stratum corneum formation and epidermal-barrier acquisition 10.

Its regulation is where the literature is densest, and it is all about getting the protein to the surface: the zymogen is glycosylated and trafficked and then autocatalytically cleaved; N-glycosylation at Asn405 and Asn445 is required for ER exit, trafficking, autocleavage, activation and secretion; phosphorylation is required for surface localisation and increases when protease activity is inhibited by HAI-2; and the cognate Kunitz inhibitors HAI-1 (SPINT1) and HAI-2 (SPINT2) are each sufficient to drive phosphorylation and surface transport, after which the protease is shed 234.

Outside its own physiology, TMPRSS13 is best known as a viral entry protease - which is the source of a query trap discussed below 11.

What the fusion joins - and why the result makes no topological sense
sidegeneexonsresiduesmodules supplied
5-primeKMT2A1-20aa 1-1888MBM 6-25, IBM1 123-134, IBM2 147-152, AT-hooks 169-309, CXXC 1147-1195
3-primeTMPRSS132-13aa 8-567 in the NM_001077263 frame (predicted, in frame)transmembrane signal-anchor, SRCR, LDL-A, peptidase S1 with its full triad

TMPRSS13 exon 2 (NM_001077263.3, nt 117-546) begins exactly 21 nucleotides after the ATG, so the junction is predicted in frame and essentially the entire protein minus its first seven residues comes with it 7. Of the seven cases this is the second-largest predicted partner contribution and by far the most functional one: an intact protease domain with an intact catalytic triad.

And that is the problem. A type II transmembrane protease reaches its substrates by being inserted into the membrane with its catalytic domain outside the cell, and its whole documented regulation - ER exit, glycosylation, phosphorylation, HAI-driven surface transport, shedding - is a secretory-pathway itinerary. KMT2A's exons 1-20 encode a nuclear module whose job is to find unmethylated CpG islands. A chimera with a signal-anchor 1,888 residues into the protein has no obvious way to be a membrane protein or a nuclear one: its signal-anchor is buried in the middle of a large cytosolically initiated polypeptide, which is a recognised route to mistargeting rather than to either destination. Nobody has looked, and the honest statement is that the chimera's localisation is unpredicted and unmeasured.

Isoform caution. The transcript the source cites, NM_001077263.3, encodes 567 aa, while UniProt's canonical isoform 1 is 586 aa - a 19-residue offset. All the domain and active-site coordinates above are on the 586 frame 1. Nineteen residues is small enough to look like a rounding error and large enough to move an active-site residue, so the partner-side residue numbers here are given in the frame they were computed in and are not mapped onto the UniProt features 7. The same trap sits on C2CD3 in this beat at 390 residues.

One more thing about this specific row: case S17 is one of the two rows in which the source table misspells the index gene as KTM2A, which this project's collision registry records at row 109. A reader grepping the table for KMT2A loses this case silently 5.

Epigenetics (PI dimension 3) - NOT APPLICABLE, and here that is the finding

The brief warned against defaulting to not-applicable, and five of these seven genes turned out to have real epigenetic content. This one does not, and the negative was earned rather than assumed.

Measured 2026-09-22, one [tiab] per term, querytranslation read back:

queryrecords
TMPRSS13[tiab] AND histone[tiab]0
TMPRSS13[tiab] AND chromatin[tiab]0
TMPRSS13[tiab] AND methylation[tiab]0
TMPRSS13[tiab] AND epigenetic[tiab]1
MSPL[tiab] AND histone[tiab] (the high-recall alias)0
control KMT2A[tiab] AND histone[tiab]218

The single epigenetic hit is not an epigenetics result: it is a whole-exome study of nine monozygotic discordant twin pairs in which the only apparent de novo variant detected by the de-novo-probability tool was in TMPRSS13 - and Sanger sequencing failed to confirm it, with the authors concluding that somatic coding mutations do not play a major role in the discordance 12. The word epigenetic appears in the abstract's framing of mutational classes. Recording this is the point of the bounded-rejection rule: a future beat re-running this query will see a 1 and should not have to re-read the paper.

And the architecture explains the zero rather than merely accompanying it. TMPRSS13 has a signal-anchor transmembrane helix, an SRCR domain, an LDL-receptor class A module and a peptidase S1 fold; it is cell-surface and secreted; its trafficking is a secretory-pathway itinerary; and its two named substrates are extracellular proproteins. There is no SET, JmjC, HAT, HDAC, bromodomain, chromodomain, Tudor, PHD or CXXC module, no nuclear-localisation route, and no non-histone protein-modification chemistry of the acyl- or methyl-transfer kind - proteolysis is not a mark. This is the gene-supt6h form of the statement, justified on domain architecture and compartment together.

Localization, cell type and cell state (PI dimensions 1 and 2)

Subcellular: cell membrane, secreted, and a cytoplasmic pool - and the UniProt record is a good one. Cell membrane carries three ECO:0000269s (PMIDs 28710277, 34562451, 35796294); Secreted one (35796294); Cytoplasm one (28710277). Rule 91 passes, and unusually the cited papers are exactly about the annotation: the 2017 paper's stated purpose is "the activation, inhibition, post-translational modification, and localization of TMPRSS13" and it reports cell-surface labelling showing the dominant surface form is phosphorylated while intracellular TMPRSS13 is predominantly non-phosphorylated 2. The topology line single-pass type II membrane protein is ECO:0000255 - a sequence-analysis rule, not an observation - which matters for the chimera argument above, because the topology assignment being rule-based is precisely what makes a buried signal-anchor's fate unpredictable.

This is the only one of the seven partners whose localisation is a surface rather than an intracellular compartment, and therefore the one whose fusion poses the sharpest compartment conflict.

Subnuclear: not applicable, correctly. No nuclear pool of any kind is reported.

Cell type: epithelial. UniProt carries three tissue comments, all ECO:0000269: placenta; and predominantly lung, placenta, pancreas and prostate, weakly in testis and peripheral-blood lymphocytes 1. In mouse the functional site is the epidermis, where deficiency impairs stratum corneum formation 10. Nothing is mesenchymal - and case S17 is a high-grade endometrial stromal sarcoma, a mesenchymal tumour of the uterus, for which no TMPRSS13 expression measurement exists.

Cell state: barrier-forming epithelium, and virus-infected cell. The epidermal-barrier phenotype is a differentiation state 10; the other well-documented state is a cell in which TMPRSS13 primes a viral fusion glycoprotein for entry 11.

Analogous biology in other fusion families (PI dimension 5) - and the worst token trap in this beat

TMPRSS13 is not a fusion partner in any tumour, and the query that seems to show otherwise is the most convincingly wrong in this beat. TMPRSS13[tiab] AND fusion[tiab] returns 9 records - and all nine are about viral membrane-fusion glycoproteins: influenza haemagglutinin, SARS-CoV-2 spike, human parainfluenza virus 3 F protein, swine acute diarrhoea syndrome coronavirus, plus an IL4I1/TMPRSS13 competition paper 11. TMPRSS13[tiab] AND translocation[tiab] = 0. AND KMT2A[tiab] = 0. AND MLL[tiab] = 0 9.

Why this one is worse than the CPT1A version of the same trap. For CPT1A, fusion means mitochondrial fusion - a different sense of an ordinary word. For TMPRSS13 the protein's actual, central, published function is cleaving proteins that are called fusion proteins. The word is not a coincidence; it is the subject matter. A beat writing "TMPRSS13 appears in nine fusion papers" would be making a true statement about the records and a false one about biology, and no amount of query hygiene would catch it - only title-checking would.

What the partner could plausibly bring, if the chimera exists: TMPRSS13 activates pro-HGF 2, and HGF/MET signalling is active in several soft-tissue sarcomas. That is the one route by which a protease-carrying chimera could have a proliferative consequence, and it requires the protease to be catalytically competent, correctly oriented and at a membrane - none of which the fusion's predicted architecture supports. It is a hypothesis-shaped remark, not a finding.

Status: what this case is and is not
  • One patient. Case S17, high-grade endometrial stromal sarcoma; one of the two rows where the source misspells the index gene as KTM2A 5.
  • The partner is predicted in frame and nearly complete, which makes this the case where the chimera would be strangest and where nobody has looked 7.
  • No fusion protein, no transcript sequence, no reciprocal product, no functional assay.
  • DepMap: nothing. Mean Chronos -0.095 in 11 KMT2A-rearranged lines against -0.029 in 43 other leukaemia lines, difference -0.067, pan-essentiality -0.035 over 1,208 lines, selectivity rank 2,409 of 17,728 13. The rank looks high only because near-zero differences sort unstably at a near-zero floor.
  • Not a treatment rationale. See risk-a-menin-inhibitor-prescribed-off-a-kmt2a-gene-name-in-a-solid-tumour.
Facts and figures
uniprotQ9BYE2
gene symbolTMPRSS13
length AA586
accession resolutiongene_exact:TMPRSS13 AND organism_id:9606 AND reviewed:true returned exactly one row, primary gene name TMPRSS13, entry TMPSD_HUMAN, canonical length 586 aa. Primary name read back and length asserted - but see isoform_frame_trap and symbol_traps: the asserted length is not the cited transcript's, and one of the record's own gene synonyms points at a different gene family.
refseq transcriptNM_001077263.3 / NP_001070731.1 (13 exons, CDS 96..1799, 567 aa; exon 2 = nt 117-546 = aa 8-151 IN THAT FRAME)
isoform frame trapUniProt canonical isoform 1 = 586 aa; the transcript the fusion source cites, NM_001077263.3 / NP_001070731.1, encodes 567 aa - a 19-residue offset, with five isoforms listed on the record. All domain and active-site coordinates here are on the 586 frame and the partner-side breakpoint residues are in the 567 frame; THE TWO ARE NOT MAPPED ONTO EACH OTHER. Nineteen residues is small enough to look like a rounding error and large enough to move an active-site residue, which makes this the more dangerous version of the 390-residue C2CD3 trap in the same beat. verify-accessions.py asserts accession plus length and so catches a wrong GENE, not a wrong ISOFORM.
symbol trapsTWO, BOTH MEASURED 2026-09-22. (1) THE APPROVED SYMBOL IS THE LOW-RECALL ONE: TMPRSS13[tiab] = 38 against MSPL[tiab] = 113 - the same inversion this project's registry records for SEPT*/SEPTIN* - BUT MSPL is also a contaminated token, with only 19 of its 113 records co-occurring with 'protease', so it must be title-checked rather than ORed in blindly (the recall-versus-precision pair recorded for RD/NELFE and U19/EAF2). (2) UniProt lists TMPRSS11 among this record's GENE SYNONYMS, and TMPRSS11 is NOT a standalone HGNC symbol - the approved symbols are TMPRSS11A/B/D/E/F, all different genes - so expanding that curated synonym reaches the wrong family, which is the rule-63 shape (a curated alias pointing elsewhere). TMPRSS11[tiab] = 2. The third synonym, MSP, is unusable.
domain coordinate frameUniProt Q9BYE2 canonical isoform 1, 586 aa - NOT the 567-aa NM_001077263 frame used for the breakpoint
epigenetic readoutNOT APPLICABLE - AND THIS IS THE ONE GENE OF THE SEVEN WHERE THAT SURVIVED BEING PROBED. The brief warned against defaulting to not-applicable, and five of these seven partners turned out to have real epigenetic content in one direction or another; this one does not, in either. BOUNDED 2026-09-22, one [tiab] per term with querytranslation read back: TMPRSS13[tiab] AND histone[tiab] = 0; AND chromatin[tiab] = 0; AND methylation[tiab] = 0; AND epigenetic[tiab] = 1; and the HIGH-RECALL ALIAS was searched too, MSPL[tiab] AND histone[tiab] = 0. Control KMT2A[tiab] AND histone[tiab] = 218 through the identical shape. The single 'epigenetic' record is PMID 26681452, a whole-exome study of nine monozygotic discordant twin pairs whose only apparent de novo variant was in TMPRSS13 AND FAILED SANGER CONFIRMATION - recorded here as a bounded rejection so a later beat does not have to re-read it. JUSTIFIED ON ARCHITECTURE AND COMPARTMENT TOGETHER, in the gene-supt6h form: TMPRSS13 is a signal-anchored, cell-surface and secreted peptidase S1 protease with an SRCR domain and an LDL-receptor class A module; it has no SET, JmjC, HAT, HDAC, bromodomain, chromodomain, Tudor, PHD or CXXC module; it has no reported nuclear pool; its named substrates are EXTRACELLULAR PROPROTEINS (pro-PRSS8/prostasin, pro-HGF); and proteolysis is not a mark, so it is not a writer, eraser or reader in any sense. THE LIMIT ON THE NEGATIVE: titles and abstracts only, supporting figures not searched (rule 80); and for the FUSION the question is genuinely open, because the chimera would place a protease domain in an unpredicted compartment and nobody has looked.
non histone substratepro-PRSS8/prostasin and pro-HGF - both cleaved as PROPROTEINS OUTSIDE THE CELL. Proteolysis is not a post-translational mark in the writer/eraser/reader sense this graph uses, and recording it under this key would be a category error: the substrate is consumed and converted, not tagged.
subnuclear localizationNOT APPLICABLE, correctly: no nuclear pool of TMPRSS13 is reported in any system. TMPRSS13 is a membrane-anchored and shed surface protease, and the meaningful distribution for it is the balance between membrane-anchored and shed forms set by stem-region cleavage kinetics, not a compartment.
cell typeEPITHELIAL. UniProt carries three ECO:0000269 tissue comments: placenta, and predominantly LUNG, PLACENTA, PANCREAS AND PROSTATE with weak expression in testis and peripheral-blood lymphocytes. In mouse the functional site is the EPIDERMIS, where deficiency impairs stratum corneum formation and epidermal barrier acquisition. NOTHING IS MESENCHYMAL - and case S17 is a high-grade ENDOMETRIAL STROMAL sarcoma, a mesenchymal uterine tumour, for which no TMPRSS13 expression measurement exists.
cell stateBARRIER-FORMING EPITHELIUM - the epidermal stratum-corneum phenotype is a differentiation state - and VIRUS-INFECTED CELL, where TMPRSS13 primes a viral fusion glycoprotein for entry. A third, subtler state is set by INHIBITOR AVAILABILITY: surface residency requires HAI-1 or HAI-2, so whether the protease is at the surface at all depends on the cell's inhibitor expression rather than on protease levels.
localization shown for fusionNO - 'nobody has looked in any fusion' (rule 86), and for this partner the question is unusually interesting because the predicted chimera is topologically incoherent. TMPRSS13[tiab] AND KMT2A[tiab] = 0 and AND MLL[tiab] = 0; no sarcoma KMT2A fusion protein has been visualised by any method.
fusion partner in other tumoursNONE REPORTED - AND THIS IS THE WORST TOKEN TRAP IN THE BEAT. TMPRSS13[tiab] AND fusion[tiab] returns 9, ALL title-checked and ALL about VIRAL MEMBRANE-FUSION GLYCOPROTEINS: influenza haemagglutinin (PMIDs 20219906, 19565019), SARS-CoV-2 spike (35913162, 36254158, 41408854), HPIV3 F protein (39382296), SADS-CoV (38808555), an IL4I1/TMPRSS13 spike-competition paper (36131918) and the epidermal-barrier paper (24832573). AND translocation[tiab] = 0. WHY IT IS WORSE THAN THE CPT1A VERSION: for CPT1A, 'fusion' is a different sense of an ordinary word (mitochondrial fusion); for TMPRSS13 the protein's central published function IS cleaving proteins that are called fusion proteins, so the word is the subject matter and no query hygiene catches it - only title-checking does. Controls: EWSR1[tiab] AND fusion[tiab] = 1,208; AND translocation[tiab] = 462. WHAT THE PARTNER COULD PLAUSIBLY BRING, LABELLED AS A HYPOTHESIS AND NOT A FINDING: TMPRSS13 activates pro-HGF, and HGF/MET signalling is active in several soft-tissue sarcomas - but that requires a catalytically competent, correctly oriented, membrane-resident protease, which the chimera's predicted architecture does not support.
absence classnever-measured
domains
Transmembrane signal-anch…166.0Transmembrane signal-anchor (type II): 166SRCR (scavenger receptor …195.0SRCR (scavenger receptor cysteine-rich): 195LDL-receptor class A204.0LDL-receptor class A: 204Peptidase S1 (trypsin-lik…326.0Peptidase S1 (trypsin-like serine protease): 326Active site (charge relay…366.0Active site (charge relay, His366): 366Active site (charge relay…414.0Active site (charge relay, Asp414): 414Active site (charge relay…511.0Active site (charge relay, Ser511): 511
aliases and retired symbols searched
[
  "TMPRSS13",
  "MSPL",
  "MSP",
  "TMPRSS11 (rejected - different gene family)"
]
domains — the numbers behind the chart
[
  {
    "name": "Transmembrane signal-anchor (type II)",
    "start": 166,
    "end": 186
  },
  {
    "name": "SRCR (scavenger receptor cysteine-rich)",
    "start": 195,
    "end": 325
  },
  {
    "name": "LDL-receptor class A",
    "start": 204,
    "end": 226
  },
  {
    "name": "Peptidase S1 (trypsin-like serine protease)",
    "start": 326,
    "end": 559
  },
  {
    "name": "Active site (charge relay, His366)",
    "start": 366,
    "end": 366
  },
  {
    "name": "Active site (charge relay, Asp414)",
    "start": 414,
    "end": 414
  },
  {
    "name": "Active site (charge relay, Ser511)",
    "start": 511,
    "end": 511
  }
]
functions
[
  "type II transmembrane serine protease; autocatalytic zymogen activation",
  "cleaves pro-PRSS8/prostasin to the active protease",
  "cleaves pro-HGF to the active growth factor, promoting MAPK signalling",
  "stratum corneum formation and epidermal barrier acquisition (by similarity to mouse)",
  "priming of viral membrane-fusion glycoproteins - influenza haemagglutinin, SARS-CoV-2 spike, HPIV3 F, SADS-CoV"
]
sarcoma case
{
  "patient": "S17",
  "diagnosis": "high-grade endometrial stromal sarcoma",
  "fusion": "KMT2A exon 20 :: TMPRSS13 exon 2 (NM_001077263)",
  "source_typo": "the source table writes KTM2A for this row - one of two such rows in eighteen; KTM2A is not an HGNC symbol and a grep for KMT2A loses the case silently (this project's collision registry, row 109)",
  "kmt2a_polarity": "5-prime - the LEUKAEMIC orientation, opposite to the recurrent sarcoma arms",
  "kmt2a_exons_retained": "1-20",
  "kmt2a_residues_retained": "aa 1-1888",
  "kmt2a_modules_retained": [
    "MBM 6-25",
    "IBM1 123-134",
    "IBM2 147-152",
    "AT-hooks 169-309",
    "CXXC 1147-1195"
  ],
  "mbm_retained": true,
  "cxxc_retained": true,
  "in_leukaemic_breakpoint_cluster_region": false,
  "partner_residues_predicted": "aa 8-567 in the NM_001077263 frame - essentially the whole protein minus its first seven residues",
  "partner_modules_predicted_retained": [
    "transmembrane signal-anchor",
    "SRCR",
    "LDL-receptor class A",
    "peptidase S1 with the complete charge-relay triad"
  ],
  "predicted_frame": "IN FRAME (TMPRSS13 exon 2 starts 21 nt after the ATG, 21 % 3 == 0; KMT2A exon 20 ends at a complete codon)",
  "frame_derivation": "computed 2026-09-22 from NM_001197104.2 and NM_001077263.3 exon coordinates - a derivation, not an observation",
  "why_the_in_frame_prediction_is_not_reassuring": "The predicted product is TOPOLOGICALLY INCOHERENT. A type II transmembrane serine protease reaches its substrates by membrane insertion with its catalytic domain outside the cell, and its entire documented regulation is a secretory-pathway itinerary: ER exit requiring N-glycosylation at Asn405 and Asn445, phosphorylation for surface residency, HAI-1/HAI-2-driven surface transport, then shedding. In the chimera the signal-anchor would sit 1,888 residues into a cytosolically initiated polypeptide behind a nuclear CpG-island-reading module. That is a recognised route to MISTARGETING rather than to either destination, and 'in frame' must not be read as 'functional protease'. Nobody has looked.",
  "cytoband_note": "TMPRSS13 is at 11q23, the KMT2A cytoband itself, so this case may be an intrachromosomal event of a kind both fusion panels and karyotyping handle poorly",
  "source": "Supplemental Table 2 of PMID 32461620",
  "fusion_protein_observed": false,
  "reciprocal_transcript_reported": false,
  "functional_assay_done": false,
  "authors_own_verdict": "of unclear significance"
}
localization
{
  "CONSTITUTIVE": "CELL MEMBRANE (three ECO:0000269s: PMIDs 28710277, 34562451, 35796294), SECRETED (one: 35796294) and a CYTOPLASMIC pool (one: 28710277). RULE 91 PASSES, and unusually well - the 2017 paper's stated purpose is 'the activation, inhibition, post-translational modification, and localization of TMPRSS13' and it reports cell-surface labelling showing the dominant surface form is phosphorylated while intracellular TMPRSS13 is predominantly non-phosphorylated. The topology line 'single-pass type II membrane protein' is ECO:0000255, a sequence-analysis rule rather than an observation - which matters here, because a rule-based topology call is exactly what makes a buried signal-anchor's fate unpredictable. NOTE: this is the ONLY one of the seven partners whose localisation is a cell SURFACE rather than an intracellular compartment, and therefore the one whose fusion poses the sharpest compartment conflict.",
  "systems": "transfected human cell lines with cell-surface labelling; mouse epidermis. No mesenchymal, sarcoma or haematopoietic system."
}
depmap kmt2a selectivity
{
  "chronos_mean_kmt2a_rearranged": -0.0949,
  "chronos_mean_other_leukaemia": -0.0285,
  "difference": -0.0665,
  "pan_essentiality_all_lines": -0.0346,
  "n_all_lines": 1208,
  "selectivity_rank": 2409,
  "rank_denominator": 17728,
  "provenance": "Joined 2026-09-22 from analyses/depmap-kmt2a-selective-scan-full/results.json (generated 2026-09-19T10:12:06Z), this project's own genome-wide scan over 11 KMT2A-rearranged and 43 other leukaemia DepMap lines. NOT re-fetched and NOT independent of that scan: every caveat it carries applies, including that the 11 KMT2A-rearranged lines are drawn from nine individuals and that the arm is myeloid-enriched. The scan records a generation timestamp and NOT a DepMap RELEASE, which is rule 83 exposure carried forward rather than fixed here.",
  "how_to_read": "THE RANK OF 2,409 IS AN ARTEFACT OF THE REGIME, NOT A SIGNAL: every value here is within 0.1 of zero, and at a near-zero floor small differences sort unstably - the z-is-not-an-effect-size problem this project documents, where places gained in a ranking correlate with the narrowness of the comparator arm rather than with effect size. A second, decisive blindness: TMPRSS13 is expressed predominantly in lung, placenta, pancreas and prostate epithelium, so a leukaemia cell-line panel is knocking out a gene most of those lines do not express, and a knockout of an unexpressed gene has no phenotype by construction. A rank is not an effect size and a difference is not a dependency (this graph's DepMap rules). Check pan_essentiality_all_lines before reading anything as selective, and note the anti-selective half is 50% of any gene set by construction, with this graph's own base rate at 45.3% - so a tally of genes in that half is the null (rule 84). The decisive limitation here is lineage: the screen is leukaemia lines, and for a sarcoma fusion that is the wrong tissue. Cellosaurus holds 592 KMT2A-commented cell lines and ZERO are sarcoma or any mesenchymal tumour, so no cell-line screen can address these cases at all."
}
queries run 2026 09 22
[
  "TMPRSS13[tiab] -> 38; MSPL[tiab] -> 113; MSPL[tiab] AND protease[tiab] -> 19; TMPRSS11[tiab] -> 2",
  "TMPRSS13[tiab] AND fusion[tiab] -> 9, ALL title-checked and ALL viral membrane-fusion papers",
  "TMPRSS13[tiab] AND translocation[tiab] -> 0",
  "TMPRSS13[tiab] AND KMT2A[tiab] -> 0",
  "TMPRSS13[tiab] AND MLL[tiab] -> 0",
  "TMPRSS13[tiab] AND histone[tiab] -> 0",
  "TMPRSS13[tiab] AND chromatin[tiab] -> 0",
  "TMPRSS13[tiab] AND methylation[tiab] -> 0",
  "TMPRSS13[tiab] AND epigenetic[tiab] -> 1 (PMID 26681452, twin WES, variant unconfirmed - not epigenetics)",
  "MSPL[tiab] AND histone[tiab] -> 0",
  "TMPRSS13[tiab] AND cancer[tiab] -> 9",
  "controls: EWSR1[tiab] AND fusion[tiab] -> 1208; EWSR1[tiab] AND translocation[tiab] -> 462; AFF1[tiab] AND KMT2A[tiab] -> 104; MLLT3[tiab] AND MLL[tiab] -> 84; KMT2A[tiab] AND histone[tiab] -> 218"
]
graph neighbours already present
[
  "gene-kmt2a",
  "gene-supt6h",
  "disease-kmt2a-r-sarcoma",
  "concept-kmt2a-fusion-polarity-is-inverted-in-sarcoma"
]
How this could bite us back
  • This entry must not be read as a menin-inhibitor rationale. Menin is a TUMOUR SUPPRESSOR in mesenchyme, and the only non-haematologic menin trial with a result, NCT05731947, was terminated by the sponsor after phase 1b for not meeting pre-defined efficacy signals. An MBM-retaining KMT2A fusion in a sarcoma is a test of a hypothesis, not an indication.
  • If menin inhibition were tried here off the gene name alone, the predicted direction of effect is not obviously anti-tumour: displacing menin from chromatin in a mesenchymal cell removes a tumour suppressor from its targets. See risk-a-menin-inhibitor-prescribed-off-a-kmt2a-gene-name-in-a-solid-tumour.
  • Every quantitative statement about the fusion in this entry is a coordinate derivation from transcript records. If the panel's exon numbering differs from RefSeq's, the frame prediction, the retained-residue count and the MBM claim all move together - one systematic error, not several independent ones.
  • A protease is a tempting target and this is not a case for one. TMPRSS13's surface activity is set by HAI-1/HAI-2 availability rather than by its own expression, its family members are partially redundant in the epidermis, and the human loss-of-function phenotype (by similarity from mouse) is an epidermal barrier defect - so a TTSP inhibitor would have a broad and poorly predicted on-target profile.
  • The protein's best-known activity is priming viral entry, so systemic inhibition of TMPRSS13 would plausibly alter host susceptibility to influenza and coronaviruses - an off-target consequence that would not appear in any oncology safety model.
  • The in-frame prediction here is the least reassuring of the seven, not the most: it predicts a topologically incoherent protein. Any therapeutic reasoning that treats it as a functional membrane protease would be building on a derivation from an exon number in a supplementary table, in a case whose row misspells the index gene.
Still open
  1. Where would a KMT2A-TMPRSS13 chimera go? Its signal-anchor would sit 1,888 residues into a cytosolically initiated polypeptide behind a nuclear CpG-island reader, which is a recognised route to mistargeting rather than to either compartment - and it is answerable with one transfection and an immunofluorescence.
  2. Is this case an intrachromosomal artefact? TMPRSS13 is at 11q23, the KMT2A cytoband itself, so a local rearrangement could produce this junction without any partner selection - and that possibility applies to no other case in the cohort.
  3. Does the retained KMT2A 1-1888 module bind menin? Exon 20 is the second-latest breakpoint of the seven, so this case tests MBM retention with the largest KMT2A contribution and without the leukaemic breakpoint-cluster-region argument.
  4. Would the chimeric protease be active at all? Surface residency of TMPRSS13 requires HAI-1 or HAI-2 and N-glycosylation at Asn405 and Asn445 - so even an in-frame, correctly folded protease domain could be inactive for want of its trafficking partners, which is testable in a cell line lacking SPINT1 and SPINT2.
  5. Is MSPL a separate entity from TMPRSS13 or the same protein under two names? The virology literature treats them as two proteases in the same titles, UniProt lists MSPL as a synonym of TMPRSS13, and the alias carries three times the recall of the approved symbol - which means the field and the database disagree about what the gene is called and possibly about what it is.
Every source behind this entry (13)
  1. E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database

    TMPRSS13 (Q9BYE2, canonical isoform 586 aa) is a type II transmembrane serine protease with a signal-anchor transmembrane helix at 166-186, an SRCR domain at 195-325 containing an LDL-receptor class A module at 204-226, a peptidase S1 domain at 326-559 with a charge-relay triad at residues 366, 414 and 511, and an N-terminal low-complexity repeat array at 9-93; it is expressed predominantly in lung, placenta, pancreas and prostate and weakly in testis and peripheral-blood lymphocytes.

    What this evidence cannot show

    Accession resolved this cycle by gene_exact query with gene_primary read back: exactly one reviewed human record, primary symbol TMPRSS13, entry TMPSD_HUMAN, canonical length 586 aa (rules 62, 65, 66). TWO THINGS TO CARRY FORWARD. (1) THE ASSERTED LENGTH IS NOT THE LENGTH OF THE TRANSCRIPT THE FUSION SOURCE CITES: NM_001077263.3 / NP_001070731.1 encodes 567 aa, 19 residues shorter, and UniProt lists five isoforms; all coordinates in this claim are on the 586 frame. (2) UniProt lists TMPRSS11 among this record's GENE SYNONYMS, and TMPRSS11 is not a standalone HGNC symbol - the approved symbols are TMPRSS11A, TMPRSS11B, TMPRSS11D, TMPRSS11E and TMPRSS11F, all different genes - so ORing that synonym into a query reaches the wrong family. The topology annotation is ECO:0000255, a sequence-analysis rule. Tissue comments are ECO:0000269 but derive largely from one 2001 expression survey.

    Evidence that argues against this (1)
    • TTSP family members share domain architecture, substrate repertoires and inhibitors, and much of the functional literature treats MSPL and TMPRSS13 together or as variants - so attributing a TTSP result specifically to TMPRSS13 rather than to a family member requires checking which construct was used. This is the same hazard this graph records for CK2 catalytic subunits, where the chemistry cannot separate the proteins. Okumura Y et al. 2010
  2. E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    TMPRSS13 is a glycosylated, active protease whose own proteolytic activity mediates zymogen cleavage; its cell-surface expression requires the Kunitz-type inhibitors HAI-1 (SPINT1) or HAI-2 (SPINT2), whose presence drives phosphorylation of the protease's intracellular domain and coincides with efficient transport to the surface and subsequent shedding, with the dominant surface form phosphorylated and the intracellular form predominantly non-phosphorylated; TMPRSS13 also cleaves pro-hepatocyte growth factor to its active form, promoting MAPK signalling.

    PMID 28710277DOI
    What this evidence cannot show

    Cell-line biochemistry with transfected constructs and cell-surface labelling; not a tumour, not a sarcoma and not a haematopoietic system. The pro-HGF activation result is from a separate paper (PMID 20977675) and is asserted here alongside because both underpin UniProt's FUNCTION comment. Abstract read from the eutils record; full text not read. RULE 91 PASSES: this is the principal ECO:0000269 citation behind the 'Cell membrane' and 'Cytoplasm' localisation comments, and the paper's stated purpose is the activation, inhibition, post-translational modification AND LOCALIZATION of TMPRSS13 - it is genuinely about the annotation.

    Evidence that argues against this (1)
    • The inhibitor-dependence result is counter-intuitive and load-bearing: a protease that needs its own Kunitz inhibitor in order to reach the surface means that surface activity is set by inhibitor availability rather than by protease expression, so measuring TMPRSS13 levels in a tumour would not tell you whether the protease is active. It also means an overexpression system without HAI-1/2 would see a protein stuck inside the cell - which is exactly the condition a KMT2A fusion construct would create. Martin CE et al. 2021
  3. E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    N-glycosylation of TMPRSS13 at Asn405 and Asn445 is required for exit from the endoplasmic reticulum and trafficking to the cell surface and for autocleavage, activation and secretion of the mature protein; TMPRSS13 cleaves the proform of PRSS8/prostasin to generate the active protein; and the SPINT2/HAI-2 interaction promotes its phosphorylation and cell-membrane localisation.

    PMID 34562451DOI
    What this evidence cannot show

    Cell-line biochemistry with site-directed mutants; not a tumour system. These are the post-translational-modification and substrate annotations UniProt carries with ECO:0000269, and rule 91 passes. Relevant to this beat because the itinerary described - ER exit, glycosylation, phosphorylation, surface transport, shedding - is a SECRETORY-PATHWAY route that a chimera initiating 1,888 residues upstream of the signal-anchor has no established way to take. Abstract read; full text not read.

    Evidence that argues against this (1)
    • All of this trafficking biology is established on constructs beginning at TMPRSS13's own N-terminus. None of it predicts what happens to a signal-anchor buried in the middle of a large cytosolically initiated polypeptide, so citing it as evidence that the chimera would be membrane-targeted would be an over-extension - the correct inference is that the chimera's fate is unknown. Martin CE et al. 2022
  4. E4moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

    TMPRSS13 zymogen activation, surface localisation and shedding are regulated by proteolytic cleavage within the non-catalytic stem region, and phosphorylation increases following inhibition of protease activity by SPINT2/HAI-2.

    PMID 35796294DOI
    What this evidence cannot show

    Cell-line biochemistry; not a tumour system. Cited here for the stem-region cleavage mechanism and for UniProt's 'Secreted' localisation comment, which rests on this paper alone. Abstract read; full text not read. It matters for the fusion argument that the stem region - which in the chimera would sit immediately downstream of 1,888 residues of KMT2A - is itself a regulatory element and not inert spacer.

    Evidence that argues against this (1)
    • A protein that is both membrane-anchored and shed occupies two compartments, so 'localisation' for TMPRSS13 is a distribution set by cleavage kinetics rather than a compartment assignment - which makes any single-compartment statement about it, including this entry's, a simplification. Murray AS et al. 2017
  5. E5moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

    In case S17 of a pan-sarcoma KMT2A-rearrangement series, a high-grade endometrial stromal sarcoma carried KMT2A exon 20 fused to TMPRSS13 exon 2 (NM_001077263), one of seven cases with KMT2A as the 5-prime partner - and it is one of the two rows in which the source table misspells the index gene as KTM2A.

    PMID 32461620DOI
    What this evidence cannot show

    This is ONE patient (case S17, high-grade endometrial stromal sarcoma) in a commercial targeted RNA-panel series enriched for advanced disease. The source authors call the non-recurrent KMT2A partners of UNCLEAR SIGNIFICANCE and several may be passengers. No fusion protein has been observed for this or any other sarcoma KMT2A fusion by any method, no reciprocal transcript was reported, no functional experiment has been done, and the panel targets specific KMT2A exons, which biases which breakpoints are detectable at all. Supplemental Table 2 reports EXON NUMBERS from the panel's own annotation - the junction nucleotide, the transcript sequence and the reading frame are not given - and it misspells the index gene as KTM2A in two of eighteen rows, so a reader grepping for KMT2A silently loses cases. This graph's parse is at analyses/sarcoma-nonrecurrent-partners/results.json. Naming this case makes it auditable; it does not promote the partner to a driver and it is not a treatment rationale. THIS ROW IS THE KTM2A ROW: the source table writes 'KTM2A exon 20 to TMPRSS13 exon 2', and KTM2A is not an HGNC symbol for anything, so it resolves to nothing and a reader grepping the table for KMT2A loses this case silently. This project's collision registry records the trap at row 109.

    Evidence that argues against this (1)
    • The cohort's recurrent, morphology- and methylation-defined KMT2A sarcoma arms are YAP1-KMT2A-YAP1 and VIM::KMT2A, and those are what the paper's title and conclusions are about. The seven 5-prime non-recurrent cases are explicitly set aside by the authors as of unclear significance, and a targeted panel run in advanced disease will also detect passenger rearrangements. A single MBM-retaining case is a lead, not an entity. Massoth LR et al. 2020
  6. E6strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database

    KMT2A exons 1-20 encode aa 1-1888 and therefore retain the menin-binding motif (6-25), LEDGF IBM1 (123-134) and IBM2 (147-152), all three AT-hooks and the CXXC domain (1147-1195); the exon-20 break lies well downstream of the leukaemic breakpoint cluster region (exons 8-14).

    What this evidence cannot show

    KMT2A exon-to-residue map, re-derived this cycle from RefSeq NM_001197104.2 (CDS 22..11940, 36 exons, NP_001184033.1) and independently reproducing this graph's own ana-kmt2a-exon-to-domain-map: exon 1 = aa 1-144 (MBM 6-25, LEDGF IBM1 123-134), exon 2 = aa 145-168 (IBM2 147-152), exon 3 = aa 168-1052 (all three AT-hooks), exon 4 = 1053-1112, exons 5-6 = 1112-1212 (CXXC 1147-1195). Exon 3 ends at aa 1052, exon 8 at 1362, exon 9 at 1406, exon 15 at 1668, exon 20 at 1888, exon 22 at 1987. All six KMT2A donor exons in this cohort end at a COMPLETE CODON (phase 0), which is why KMT2A can fuse in frame to so many acceptors. Note NM_001197104.2 encodes 3,972 aa while UniProt Q03164 is 3,969 aa; the three-residue difference lies downstream of every coordinate used here. This is a COORDINATE DERIVATION from a transcript record, not a measurement of a fusion protein: no sarcoma KMT2A fusion protein has ever been observed by any method and no menin-binding assay has been run on one.

    Evidence that argues against this (1)
    • Retaining a motif is not the same as presenting it. The chimera's folding, stability and nuclear import are all unmeasured, and a fusion protein can retain a binding motif in primary sequence while failing to engage its partner. This graph's queued test exp-menin-pulldown-on-yky-versus-kmt2a-aff1 exists precisely because retention has never been shown to imply binding for any sarcoma arm. Massoth LR et al. 2020
  7. E7moderate confidence 3 of 5: One strong study, or several smaller consistent ones.database

    TMPRSS13 exon 2 (NM_001077263.3, nt 117-546, CDS from 96) begins exactly 21 nucleotides after the start codon, so a junction from a phase-0 KMT2A donor exon is predicted in frame and the chimera would carry TMPRSS13 from residue 8 onward - essentially the whole protein including the signal-anchor transmembrane helix, the non-catalytic stem and the complete peptidase S1 domain.

    What this evidence cannot show

    A derivation computed this cycle from two RefSeq records, not an observation: the commercial panel's exon numbering need not match RefSeq's, the junction nucleotide is unknown, and an in-frame prediction says nothing about whether the transcript is made or translated. THE ISOFORM CAUTION MATTERS HERE: NM_001077263.3 encodes 567 aa while UniProt's canonical is 586, so the partner-side residue numbers in this claim are given in the RefSeq frame and are NOT mapped onto the UniProt domain coordinates; a 19-residue offset is small enough to pass unnoticed and large enough to move an active-site residue. AND THE BIOLOGICAL CAUTION: an in-frame prediction here produces a TOPOLOGICALLY INCOHERENT protein - a signal-anchor buried 1,888 residues into a cytosolically initiated polypeptide - so 'in frame' should not be read as 'functional protease'.

    Evidence that argues against this (1)
    • An in-frame prediction is not a demonstration of a product, and the same calculation returns the opposite answer for other members of this cohort: C2CD3 exon 24 and RALGAPA2 exon 2 both begin at phase 1 and are predicted FRAMESHIFTED, and AHDC1 exon 2 is entirely 5'-untranslated. That the derivation discriminates between cases is a reason to trust the method, not any one result. Massoth LR et al. 2020
  8. E8strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database

    TMPRSS13 writes, erases and reads no histone mark and has no epigenetic content in any direction: four bounded queries return zero or noise, and its domain architecture and compartment explain the absence - a signal-anchored, cell-surface and secreted peptidase S1 protease whose named substrates are extracellular proproteins has no route to a histone, and proteolysis is not a mark.

    What this evidence cannot show

    BOUNDED 2026-09-22, one [tiab] per term with querytranslation read back (rules 94, 95): TMPRSS13[tiab] AND histone[tiab] = 0; AND chromatin[tiab] = 0; AND methylation[tiab] = 0; AND epigenetic[tiab] = 1 (PMID 26681452, see E12 - not an epigenetics result); and the HIGH-RECALL ALIAS was searched too, MSPL[tiab] AND histone[tiab] = 0. Control through the identical shape: KMT2A[tiab] AND histone[tiab] = 218. TITLES AND ABSTRACTS ONLY; supporting figures were not searched (rule 80), which for a measurement-shaped query is the standing limitation. The architecture argument bounds INTRINSIC activity: no SET, JmjC, HAT, HDAC, bromodomain, chromodomain, Tudor, PHD or CXXC module, no nuclear-localisation route reported, and no acyl- or methyl-transfer chemistry. This is the one gene of the seven where the not-applicable answer survived being probed - the brief warned against defaulting to it, and five of the seven turned out to have real epigenetic content.

    Evidence that argues against this (1)
    • A domain-architecture-plus-compartment argument is still an argument from absence of evidence, and proteases do reach the nucleus in some settings - HTRA2, and separated protease domains after shedding, are precedents for surface proteases with intracellular activities. The defensible claim is that no epigenetic activity has been reported or is architecturally expected, not that none is possible; and for the FUSION the question is entirely open because the chimera would place the protease domain in an unpredicted compartment. Massoth LR et al. 2020
  9. E9moderate confidence 3 of 5: One strong study, or several smaller consistent ones.paper

    TMPRSS13 is not a reported gene-fusion partner in any tumour, and the symbol carries two measurable identifier traps: the approved symbol has roughly a third of its alias's recall (TMPRSS13[tiab] = 38 against MSPL[tiab] = 113), and UniProt lists TMPRSS11 as a gene synonym although TMPRSS11 is not a standalone HGNC symbol.

    PMID 32461620DOI
    What this evidence cannot show

    Bounded 2026-09-22 with one [tiab] per term and querytranslation read back: TMPRSS13[tiab] AND fusion[tiab] = 9, ALL title-checked and ALL about viral membrane-fusion glycoproteins; AND translocation[tiab] = 0; AND KMT2A[tiab] = 0; AND MLL[tiab] = 0. Symbol-recall measurements the same day: TMPRSS13[tiab] = 38, MSPL[tiab] = 113, MSPL[tiab] AND protease[tiab] = 19, TMPRSS11[tiab] = 2. NOTE BOTH HALVES OF THE ALIAS PROBLEM: MSPL has higher recall AND is a contaminated token - only 19 of its 113 records co-occur with 'protease' - so it must be title-checked rather than ORed in blindly, which is the recall-versus-precision pair this project's registry records for RD/NELFE and U19/EAF2. The third synonym, MSP, is unusable (macrophage-stimulating protein and others). Positive controls through the IDENTICAL one-tag-per-term query shape, run 2026-09-22 (rule 67, rule 94): EWSR1[tiab] AND fusion[tiab] = 1,208; EWSR1[tiab] AND translocation[tiab] = 462; AFF1[tiab] AND KMT2A[tiab] = 104; MLLT3[tiab] AND MLL[tiab] = 84; KMT2A[tiab] AND histone[tiab] = 218. Every query was written with one [tiab] per term and its querytranslation read back from the esearch response, so no parenthesised group silently expanded to [All Fields]. Bounded by TITLES AND ABSTRACTS, not by papers (rule 80): supporting figures were not searched.

    Evidence that argues against this (1)
    • A literature absence cannot see an unreported rearrangement, and TMPRSS13 sits at 11q23 - the KMT2A locus's own cytoband - so a KMT2A-TMPRSS13 fusion may be an intrachromosomal event of a kind that fusion panels and karyotyping both handle poorly. That proximity is worth flagging as a reason the case could be an artefact of local rearrangement rather than a partner selection. Massoth LR et al. 2020
  10. E10strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    TMPRSS13 deficiency impairs stratum corneum formation and epidermal barrier acquisition in mouse.

    TMPRSS13 deficiency impairs stratum corneum formation and epidermal barrier acquisition. · Madsen DH, Szabo R, Molinolo AA et al · Biochemical Journal 2014
    PMID 24832573DOI
    What this evidence cannot show

    MOUSE knockout, epidermal development (rule 73: the system is in the claim). This is the loss-of-function phenotype UniProt carries by similarity for the human protein, so the human statement is ECO:0000250. Epidermis is epithelial; case S17 is a mesenchymal uterine tumour, so this phenotype is in neither the index disease's lineage nor the tumour's. Abstract read; full text not read.

    Evidence that argues against this (1)
    • A developmental barrier phenotype in mouse skin is a long way from any tumour biology, and TTSP family members are partially redundant in the epidermis - prostasin, matriptase and TMPRSS13 all act in overlapping proteolytic cascades - so a single-knockout phenotype does not establish that TMPRSS13 is the rate-limiting protease in any human tissue. Martin CE et al. 2021
  11. E11strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    TMPRSS13 and MSPL proteolytically activate the membrane-fusion activity of the haemagglutinin of highly pathogenic avian influenza viruses and support their multicycle replication, and TMPRSS13 has since been reported as an entry protease for SARS-CoV-2 spike, human parainfluenza virus 3 F protein and swine acute diarrhoea syndrome coronavirus - which is why every record returned by 'TMPRSS13 AND fusion' is about viral fusion proteins rather than gene fusions.

    PMID 20219906DOI
    What this evidence cannot show

    Virology in cell culture; nothing to do with cancer or with gene rearrangement. It is cited here for exactly one purpose: to document WHY the fusion-partner query for this gene returns a healthy-looking count of noise. All nine records returned by TMPRSS13[tiab] AND fusion[tiab] were title-checked on 2026-09-22 and all nine concern viral membrane-fusion glycoproteins. This trap is worse than the CPT1A version because the word 'fusion' here is not a different sense of an ordinary word - it is the protein's actual published subject matter, so no query hygiene catches it and only title-checking does.

    Evidence that argues against this (1)
    • The viral-entry literature is also where the MSPL/TMPRSS13 naming ambiguity is worst - this paper's own title treats MSPL and TMPRSS13 as two proteases - so attributing a specific entry-priming activity to TMPRSS13 as opposed to its variant requires checking the construct in each paper. UniProt Consortium 2026
  12. E12strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

    The single record returned by 'TMPRSS13 AND epigenetic' is a whole-exome study of nine monozygotic discordant twin pairs in which the only apparent de novo variant found by a de-novo-probability tool was in TMPRSS13 - and Sanger sequencing failed to confirm it, the authors concluding that somatic coding mutational events do not play a major role in the discordance.

    Whole-Exome Sequencing in Nine Monozygotic Discordant Twins. · Zhang R, Thiele H, Bartmann P et al · Twin Research and Human Genetics 2016
    PMID 26681452DOI
    What this evidence cannot show

    Recorded as a BOUNDED REJECTION rather than as evidence about TMPRSS13, per this project's rule that papers read and judged off-beat should leave a trace. The word 'epigenetic' appears only in the abstract's framing of mutational classes; there is no epigenetic measurement, and the TMPRSS13 variant was explicitly NOT confirmed. Nine twin pairs, congenital heart defects and other discordant phenotypes - nothing to do with cancer. A future beat re-running TMPRSS13[tiab] AND epigenetic[tiab] will see a 1 and should not have to re-read the paper.

    Evidence that argues against this (1)
    • An unconfirmed variant in a small twin study is not evidence of anything about TMPRSS13, and citing it at all risks giving it standing it does not have - which is why it is filed here as the reason a count is 1 rather than 0, and not as support for any claim. Massoth LR et al. 2020
  13. E13weak confidence 2 of 5: Preliminary, small, indirect, or not yet peer reviewed.dataset

    TMPRSS13 shows no meaningful dependency in this project's own DepMap scan: mean Chronos -0.095 in 11 KMT2A-rearranged lines against -0.029 in 43 other leukaemia lines, difference -0.067, pan-essentiality -0.035 across 1,208 lines, selectivity rank 2,409 of 17,728.

    DepMap CRISPR (Chronos) gene-effect scan of 11 KMT2A-rearranged versus 43 other leukaemia lines (this project's own scan) · Broad Institute DepMap; joined locally from analyses/depmap-kmt2a-selective-scan-full · analyses/depmap-kmt2a-selective-scan-full 2026
    What this evidence cannot show

    Joined 2026-09-22 from analyses/depmap-kmt2a-selective-scan-full/results.json (generated 2026-09-19T10:12:06Z), this project's own genome-wide scan over 11 KMT2A-rearranged and 43 other leukaemia DepMap lines. NOT re-fetched and NOT independent of that scan: every caveat it carries applies, including that the 11 KMT2A-rearranged lines are drawn from nine individuals and that the arm is myeloid-enriched. The scan records a generation timestamp and NOT a DepMap RELEASE, which is rule 83 exposure carried forward rather than fixed here. The apparently high rank of 2,409 is an artefact of the regime rather than a signal: every value here is within 0.1 of zero, and at a near-zero floor small differences sort unstably - this is the z-is-not-an-effect-size problem this project documents, where places gained in the ranking correlate with the narrowness of the comparator arm rather than with effect. A rank is not an effect size and a difference is not a dependency (this graph's DepMap rules). Check pan_essentiality_all_lines before reading anything as selective, and note the anti-selective half is 50% of any gene set by construction, with this graph's own base rate at 45.3% - so a tally of genes in that half is the null (rule 84). The decisive limitation here is lineage: the screen is leukaemia lines, and for a sarcoma fusion that is the wrong tissue. Cellosaurus holds 592 KMT2A-commented cell lines and ZERO are sarcoma or any mesenchymal tumour, so no cell-line screen can address these cases at all.

    Evidence that argues against this (1)
    • TMPRSS13 is expressed predominantly in lung, placenta, pancreas and prostate epithelium, so a leukaemia cell-line panel is testing a gene most of those lines do not express - a knockout of an unexpressed gene has no phenotype by construction, and the near-zero score carries no information about the gene's importance where it is expressed. UniProt Consortium 2026
Connected to