No fusion oncoprotein in any family has a published replication-protein-A measurement: 23 queries across five fusion families and four spelling variants each, every querytranslation read back, zero records - so the PI's analogy dimension for RPA has no comparator at all, and the index fusion's single locus-level result is the entire field
also called RPA in other fusion families · cross-family RPA absence
One of this project's standing questions is whether a finding in the leukaemia fusion we study also appears in other cancer-causing fusion proteins. For the DNA-coating clamp RPA, the answer is that there is nothing to compare with: across Ewing sarcoma, chronic myeloid leukaemia, acute promyelocytic leukaemia and two other childhood leukaemia fusions, not a single paper mentions RPA and the fusion together.
The same thing, for a scientist
**The PI's fifth dimension - the analogous biology in other fusion families - is empty for RPA, and the emptiness is unusually clean.** Twenty-three PubMed queries run 2026-09-24 through `scripts/pubmed_safe.py`, one `[tiab]` tag per term (rule 94), every `querytranslation` read back and every zero required to carry the genuine-zero signature [E1]. | fusion family | spellings tried | records naming RPA | |---|---|---:| | **EWSR1::FLI1** | `"EWS-FLI1"`, `"EWSR1-FLI1"`, `"EWS/FLI"`, `"EWS-FLI"`, plus `"Ewing sarcoma"` | **0** | | **BCR::ABL1** | `"BCR-ABL"`, `"BCR-ABL1"`, `"BCR/ABL"` | **0** | | **PML::RARA** | `"PML-RARA"`, `"PML/RARA"`, `"PML-RAR"` | **0** | | **ETV6::RUNX1** | `"ETV6-RUNX1"`, `"ETV6/RUNX1"`, `"TEL-AML1"` | **0** | | **NUP98 fusions** | `NUP98` against 'replication protein A' and against `RPA2` | **0** | | generic | `"fusion oncoprotein"` | **0** | **The operands all work**: `"replication protein A"[tiab]` = 1,569, `"BCR-ABL"[tiab]` = 10,221, `"PML-RARA"[tiab]` = 841, `NUP98[tiab]` = 863, `"EWS-FLI1"[tiab]` = 650, `"ETV6-RUNX1"[tiab]` = 603 [E1]. **BCR-ABL alone has ten thousand papers and none of them names RPA.** One query returns something and it is instructive: `"fusion protein"[tiab] AND "replication protein A"[tiab]` = **9**, and **all nine were opened and all nine use 'fusion protein' to mean a recombinant tagged construct** - GFP-RPA single-molecule reporters, GST and baculovirus constructs - **not an oncogenic fusion** [E2]. `absence_class: never-measured`.
What this absence is worth, stated carefully
A zero across five families with four spellings each, against operands of 650 to 10,221 records, is a strong bibliographic absence and a weak claim about biology. Three things bound it.
- Titles and abstracts only (rule 80). RPA2 phosphorylation and RPA2 foci are standard replication-stress readouts, which means they appear in figures and supplements of papers about something else. A Ewing sarcoma R-loop paper could show an RPA2 blot in Figure 3 and never say 'RPA' in its abstract. This is the single largest hole in the absence and it runs in the direction of under-counting.
- The subunit spellings were not run against every family.
RPA2was run against NUP98 and against Ewing sarcoma;RPA1,RPA3,RPA32andRPA34were not run against any family other than MLL/KMT2A. Given that the KMT2A-locus paper was reachable only by the complex name, and the menin papers only by the subunit symbol, a complete sweep needs both forms against each family and this one does not have it. - Absence of a named comparison is not absence of the biology. Replication stress and R-loops are extensively documented in Ewing sarcoma and in other fusion cancers; RPA is necessarily involved in all of it. What is absent is anyone naming RPA in that context - which is a statement about how the field writes, not about what the cells do.
So the honest form of this entry is: the analogy dimension cannot be filled from the literature, because no comparator has been written down. That is a real and useful result for a project whose standing instruction is to look for analogous biology - it says the analogy must be generated, not retrieved.
The one comparator that does exist is in the index family, and it is not an interaction
The entire published intersection of RPA with any fusion oncoprotein family, after this sweep, is:
- One locus-level occupancy measurement in the KMT2A family: RPA ChIP at the MLL 11q23 breakpoint cluster region after etoposide, in cells that contain no fusion 3. It is about how the translocation is made.
- One declined prey row: RPA2 at 3|3 spectra under MLL-AF4 in the Skucha AP-MS, adjudicated and declined elsewhere in this corpus.
- One menin contact, in HeLa and breast lines, with no fusion in any arm.
None of those is an RPA-fusion-protein interaction, in any family. So the comparator table this beat would have built has no rows, and the correct entry is this absence rather than an analogy node with a single speculative mapping.
One asymmetry is worth recording because it suggests where the field would look first if anyone did: the index family is the only one whose defining lesion is a chemotherapy-induced double-strand break at a mapped hotspot. EWSR1::FLI1, BCR::ABL1, PML::RARA and ETV6::RUNX1 have no comparable therapy-related-breakpoint literature with a named RPA assay. That may be why the one measurement that exists is here - it is the only family where somebody had a reason to put a primer set on the breakpoint.
| what is absent | Any published measurement, of any kind, connecting the replication protein A complex or any of its subunits to a fusion oncoprotein outside the KMT2A family - specifically EWSR1::FLI1, BCR::ABL1, PML::RARA, ETV6::RUNX1 or any NUP98 fusion - in a title or abstract. |
|---|---|
| what is NOT absent | (1) RPA biology itself - 1,569 title/abstract records. (2) Replication stress and R-loop biology in several of these fusion families, which is extensive and which necessarily involves RPA without naming it. (3) A locus-level RPA occupancy measurement at the KMT2A breakpoint cluster region (PMID 21048951) - in cells with NO fusion. (4) A single declined RPA2 prey row under MLL-AF4 in the only MLL-fusion AP-MS dataset. (5) A menin-RPA2 physical interaction, in HeLa and breast lines with no fusion present. (6) RPA2 appearing as a hub gene in an Ewing sarcoma differential-expression screen (PMID 31213834), which is expression, not mechanism, and not a fusion interaction. |
| absence class | never-measured |
| absence class note | The class is `never-measured` for the CROSS-FAMILY comparison specifically: no instrument has been pointed at RPA in an EWSR1::FLI1, BCR::ABL1, PML::RARA or ETV6::RUNX1 cell in a way anyone has written down. It is NOT `measured-not-reported`, because that class implies an assay whose output was discarded - and here the likelier truth is a third thing this taxonomy handles badly: the assay WAS run (RPA2 foci and phospho-RPA2 are the standard replication-stress readouts) and the result is buried in figures of papers about something else, unindexed. That is closest to `measured-not-searched`, and it is recorded as a competing class rather than resolved. ARCHIVE CHECK NOT RUN: no GEO, PRIDE or ProteomeXchange query was made for any of these five families, so per the standing instruction this absence is bounded by literature alone and is weaker for it. |
| competing absence class | measured-not-searched - see note above. The two classes differ by three orders of magnitude in the cost of closing, and this beat could not distinguish them without a full-text or archive sweep it did not run. |
| bounded by | PubMed titles and abstracts, 2026-09-24, one archive, one day |
| shown for fusion | NO - nobody has looked in any fusion family other than KMT2A, and in KMT2A nobody has looked at the fusion PROTEIN. This is the plain-NO case rule 86 reserves for 'nobody has looked in any fusion', and there is not even a mouse-model partial credit to record. |
name list rpa side
{
"tried_against_all_five_families": [
"\"replication protein A\""
],
"tried_against_some_families": [
"RPA2 (NUP98, Ewing sarcoma)"
],
"not_tried_against_any_family": [
"RPA1",
"RPA3",
"RPA32",
"RPA34",
"RFA2",
"REPA1",
"REPA2",
"REPA3",
"HSSB",
"RP-A",
"RF-A",
"\"RPA complex\""
],
"why_this_matters": "The one KMT2A-locus paper is reachable ONLY by the complex name and the menin literature ONLY by the subunit symbol. A one-sided name list is demonstrably capable of missing an entire literature, and this sweep is one-sided."
}name list fusion side
{
"tried": [
"\"EWS-FLI1\"",
"\"EWSR1-FLI1\"",
"\"EWS/FLI\"",
"\"EWS-FLI\"",
"\"Ewing sarcoma\"",
"\"BCR-ABL\"",
"\"BCR-ABL1\"",
"\"BCR/ABL\"",
"\"PML-RARA\"",
"\"PML/RARA\"",
"\"PML-RAR\"",
"\"ETV6-RUNX1\"",
"\"ETV6/RUNX1\"",
"\"TEL-AML1\"",
"NUP98",
"\"fusion oncoprotein\"",
"\"fusion protein\""
],
"not_tried": [
"CBFB::MYH11",
"RUNX1::RUNX1T1 / AML1-ETO",
"TCF3::PBX1",
"TCF3::HLF",
"CBFA2T3::GLIS2",
"DUX4::IGH",
"NPM1::ALK",
"ETV6::NTRK3",
"every KMT2A partner other than AF4",
"the specific NUP98 partner fusions (NUP98-NSD1, NUP98-KDM5A, NUP98-HOXA9)"
]
}queries — the numbers behind the chart
[
{
"q": "\"EWS-FLI1\"[tiab] AND \"replication protein A\"[tiab]",
"count": 0,
"genuine_zero": true
},
{
"q": "\"EWSR1-FLI1\"[tiab] AND \"replication protein A\"[tiab]",
"count": 0,
"genuine_zero": true
},
{
"q": "\"EWS/FLI\"[tiab] AND \"replication protein A\"[tiab]",
"count": 0,
"genuine_zero": true
},
{
"q": "\"EWS-FLI\"[tiab] AND \"replication protein A\"[tiab]",
"count": 0,
"genuine_zero": true
},
{
"q": "\"Ewing sarcoma\"[tiab] AND \"replication protein A\"[tiab]",
"count": 0,
"genuine_zero": true
},
{
"q": "RPA32[tiab] AND \"EWS-FLI1\"[tiab]",
"count": 0,
"genuine_zero": true
},
{
"q": "RPA[tiab] AND \"EWS-FLI1\"[tiab]",
"count": 0,
"genuine_zero": true
},
{
"q": "RPA2[tiab] AND \"Ewing sarcoma\"[tiab]",
"count": 1,
"ids": [
"31213834"
],
"on_topic": 0,
"note": "differential-expression hub-gene screen, not mechanism"
},
{
"q": "\"BCR-ABL\"[tiab] AND \"replication protein A\"[tiab]",
"count": 0,
"genuine_zero": true
},
{
"q": "\"BCR-ABL1\"[tiab] AND \"replication protein A\"[tiab]",
"count": 0,
"genuine_zero": true
},
{
"q": "\"BCR/ABL\"[tiab] AND \"replication protein A\"[tiab]",
"count": 0,
"genuine_zero": true
},
{
"q": "RPA32[tiab] AND \"BCR-ABL\"[tiab]",
"count": 0,
"genuine_zero": true
},
{
"q": "\"PML-RARA\"[tiab] AND \"replication protein A\"[tiab]",
"count": 0,
"genuine_zero": true
},
{
"q": "\"PML/RARA\"[tiab] AND \"replication protein A\"[tiab]",
"count": 0,
"genuine_zero": true
},
{
"q": "\"PML-RAR\"[tiab] AND \"replication protein A\"[tiab]",
"count": 0,
"genuine_zero": true
},
{
"q": "\"ETV6-RUNX1\"[tiab] AND \"replication protein A\"[tiab]",
"count": 0,
"genuine_zero": true
},
{
"q": "\"ETV6/RUNX1\"[tiab] AND \"replication protein A\"[tiab]",
"count": 0,
"genuine_zero": true
},
{
"q": "\"TEL-AML1\"[tiab] AND \"replication protein A\"[tiab]",
"count": 0,
"genuine_zero": true
},
{
"q": "NUP98[tiab] AND \"replication protein A\"[tiab]",
"count": 0,
"genuine_zero": true
},
{
"q": "NUP98[tiab] AND RPA2[tiab]",
"count": 0,
"genuine_zero": true
},
{
"q": "\"fusion oncoprotein\"[tiab] AND \"replication protein A\"[tiab]",
"count": 0,
"genuine_zero": true
},
{
"q": "fusion[tiab] AND oncoprotein[tiab] AND \"replication protein A\"[tiab]",
"count": 0,
"genuine_zero": true
},
{
"q": "\"fusion protein\"[tiab] AND \"replication protein A\"[tiab]",
"count": 9,
"on_topic": 0,
"note": "all nine are recombinant tagged constructs; see E2"
}
]operand controls
{
"\"replication protein A\"[tiab]": 1569,
"\"BCR-ABL\"[tiab]": 10221,
"NUP98[tiab]": 863,
"\"PML-RARA\"[tiab]": 841,
"\"EWS-FLI1\"[tiab]": 650,
"\"ETV6-RUNX1\"[tiab]": 603,
"note": "every operand returns hundreds to thousands of records through the identical query shape, so the conjunction zeros are about the conjunctions and not about the instrument (rule 67)"
}what would close it
[ "A full-text or figure-level search rather than a title/abstract search. RPA2 foci and phospho-RPA2 are standard readouts and would be in figures - Europe PMC full-text search is the cheapest route and was not used.", "A GEO or PRIDE sweep for RPA occupancy or RPA-containing proteomics in any of the five families. Not run.", "The breakpoint-level version of the question, which is the better analogy for the one result that exists: does RPA load at the RUNX1 or PML breakpoint after etoposide, as it does at the KMT2A BCR? Not queried in any form." ]
measurements — the numbers behind the chart
[
{
"quantity": "count",
"subject": "records naming replication protein A with any of five fusion-oncoprotein families in title or abstract",
"value": 0,
"units": "records",
"n": 22,
"universe": "22 conjunction queries across EWSR1::FLI1, BCR::ABL1, PML::RARA, ETV6::RUNX1 and NUP98 with 4 spellings each on the fusion side, PubMed title/abstract, 2026-09-24, every querytranslation read back",
"null": 1569,
"note": "null is the count for \"replication protein A\"[tiab] alone. The fusion-side operands range from 603 to 10,221 records, so the zero is not a small-literature artefact on either side."
},
{
"quantity": "count",
"subject": "on-topic hits among the 9 records matching \"fusion protein\" AND \"replication protein A\"",
"value": 0,
"units": "records",
"n": 9,
"universe": "all nine titles retrieved by esummary and read, 2026-09-24",
"null": null,
"note": "all nine use 'fusion protein' for a recombinant tagged construct. The token is contaminated for this project's purposes and was NOT added to the collision registry by this beat."
},
{
"quantity": "count",
"subject": "published RPA measurements in any fusion-oncoprotein family, of any design",
"value": 1,
"units": "measurements",
"n": 6,
"universe": "the six fusion families considered by this beat (KMT2A plus the five above)",
"null": 0,
"note": "the one is PMID 21048951, a locus-occupancy measurement in cells containing NO fusion. Counting it as a cross-family comparator is generous and the entry says so."
}
]- An empty analogy dimension is an invitation to manufacture one, and that is the risk this entry creates. With no comparator in any other family, a later beat could be tempted to reason 'RPA must behave the same way in EWSR1::FLI1' - which would be inference presented as analogy. The correct use of this entry is to say the comparison has never been made.
- Does a Europe PMC FULL-TEXT search - not title/abstract - find RPA2 blots or foci in EWSR1::FLI1, BCR::ABL1 or ETV6::RUNX1 papers? This is the cheapest way to convert the absence from `never-measured` to `measured-not-searched`, and the two classes differ by orders of magnitude in what it costs to close them.
- Does RPA load at the RUNX1 or the PML breakpoint after etoposide, as it does at the KMT2A breakpoint cluster region? Both are etoposide-associated therapy-related translocation loci, and this is the breakpoint-level analogy that actually matches the one result that exists. Never queried in any form.
- Should '"fusion protein"' be added to data/symbol-collisions.json as a contaminated token? Nine of nine hits in this sweep meant a recombinant tag. This beat did not add it, and a future beat searching fusion-oncoprotein biology with that phrase will hit the same nine papers.
- E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
Twenty-three PubMed queries pairing replication protein A names with five fusion-oncoprotein families across multiple spellings each returned zero records in every case: EWSR1::FLI1 under four spellings plus 'Ewing sarcoma', BCR::ABL1 under three, PML::RARA under three, ETV6::RUNX1 under three including TEL-AML1, NUP98 against both 'replication protein A' and RPA2, and the generic 'fusion oncoprotein'; while the operands return 1,569 for "replication protein A"[tiab], 10,221 for "BCR-ABL"[tiab], 863 for NUP98[tiab], 841 for "PML-RARA"[tiab], 650 for "EWS-FLI1"[tiab] and 603 for "ETV6-RUNX1"[tiab].
What this evidence cannot show
BOUNDED BY TITLES AND ABSTRACTS (rule 80), and that bound is unusually costly here because RPA2 foci and RPA2 phosphorylation are standard replication-stress READOUTS that live in figures and supplements of papers about something else. BOUNDED BY A NAME LIST (rule 70): the RPA side used 'replication protein A' against all five families and RPA2 against only two of them; RPA1, RPA3, RPA32, RPA34, RFA2, REPA1/2/3, HSSB, 'RP-A' and 'RF-A' were NOT run against any family. This matters concretely - the one KMT2A-locus paper is reachable ONLY by the complex name and the menin papers ONLY by the subunit symbol, so a one-sided name list is demonstrably capable of missing a whole literature. Fusion side: five families; NOT tried were CBFB::MYH11, RUNX1::RUNX1T1, TCF3::PBX1, TCF3::HLF, CBFA2T3::GLIS2, DUX4::IGH, NPM1::ALK, ETV6::NTRK3 and every KMT2A partner other than AF4. Single archive: PubMed. No GEO, PRIDE or ProteomeXchange query is part of this item.
Evidence that argues against this (1)
- The counter-case is that this absence is measuring a writing convention rather than a scientific gap. Replication stress, R-loops and fork collapse are well documented in several of these families - Ewing sarcoma most of all - and RPA is mechanistically unavoidable in all of it. A paper reporting 'increased ssDNA' or 'increased replication stress' in EWSR1::FLI1 cells almost certainly measured RPA and simply did not put the name in its abstract. The absence is therefore genuine at the level of 'nobody has written down an RPA-versus-fusion comparison' and NOT at the level of 'nobody has measured RPA in these cells'. OncoTargets and Therapy 2019
- E2established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
The only non-zero conjunction in this sweep is "fusion protein"[tiab] AND "replication protein A"[tiab], which returns 9 records, and all nine were title-checked and all nine use 'fusion protein' to mean a recombinant tagged construct rather than an oncogenic fusion: single-molecule fluorescent-RPA reporters (PMIDs 25697222, 24625741, 21225324), a RHINO/9-1-1 stoichiometry study (25602520), a rice OsEXO1 interaction study (18231866), a yeast telomere-senescence study (17202155), the RBT1 co-activator paper (10982866), a baculovirus XPA overexpression and purification paper (10833384), and a papovavirus E1/large-T RPA-interaction paper (10233951).
What this evidence cannot show
Titles only - the nine abstracts were not read, so 'all nine are recombinant-construct usages' is a judgement from titles plus journal plus year, not from full text. For six of the nine (the single-molecule, purification and virology papers) the reading is unambiguous from the title alone; for PMIDs 25602520, 18231866 and 17202155 it rests on the fact that no oncogenic fusion exists in those systems (human checkpoint clamp, rice, yeast). This item is a control on a token, not a literature review.
Evidence that argues against this (1)
- This control demonstrates the opposite of what a reader might take from it: the token 'fusion protein' is CONTAMINATED for this project's purposes, in the same family as the 'speckle' and 'exo1' rows in this corpus's collision registry, and any future beat searching for fusion-oncoprotein biology with that phrase will get nine recombinant-tag papers and no signal. That is worth recording as a token hazard in its own right, and it was not added to the registry by this beat - which is a gap this entry names rather than hides. Nucleic Acids Research 2000
- E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
The only published measurement connecting replication protein A to any fusion-oncoprotein family is a locus-level occupancy result in the KMT2A family: RPA ChIP enrichment at the MLL 11q23 breakpoint cluster region after etoposide, transient in ATM-proficient cells and spatially extended and persistent in ATM-deficient cells - measured in an SV40-transformed fibroblast pair and a CML-derived B-cell line, none of which carries a KMT2A fusion.
ATM modulates the loading of recombination proteins onto a chromosomal translocation breakpoint hotspot. · Sun J, Oma Y, Harata M, Kono K, Shima H, Kinomura A, et al. · PLoS One 5(10):e13554 2010What this evidence cannot show
Full text read from PMC 2026-09-24. It is a measurement at the LOCUS, not a measurement of the fusion protein, and no arm of the experiment contains a fusion - so calling it 'the comparator in the index family' is already generous, and it cannot anchor an analogy to another family's fusion PROTEIN. Candidate-locus ChIP-qPCR with one control region; single group; little follow-up.
Evidence that argues against this (1)
- Because this result is about the breakpoint rather than the fusion, the natural cross-family comparison is NOT to other fusion proteins but to other recurrent breakpoints - and that comparison was not attempted here. Therapy-related translocations involving RUNX1 and PML are also etoposide-associated, so 'does RPA load at the RUNX1 or PML breakpoint after etoposide' is the analogous question, and this beat did not query it. That is a named, unfollowed thread and arguably the better version of the analogy dimension for this beat. NCBI PubMed E-utilities 2026
the cross-family analogy dimension for this beat, recorded as a bounded absence
- related to →KMT2A::AFF1 (MLL-AF4), t(4;11)(q21;q23)
the index fusion is the only fusion family with any RPA measurement at all, and it is a locus measurement in cells without a fusion
- related to →ETV6::RUNX1 (TEL-AML1), t(12;21)(p13;q22)
zero records name replication protein A with this fusion under three spellings
10,221 BCR-ABL records and none of them names replication protein A
- related to →PML::RARA, t(15;17)(q24;q21)
zero records under three spellings