The rule-64 complex-form query that the RPA2 decline flagged as unrun has now been run, and it finds a published damage-state RPA occupancy measurement AT the KMT2A breakpoint cluster region - so two of this corpus's absence entries carry ids that are broader than the claims inside them
also called rule-64 complex-form query for RPA versus KMT2A · Q192 complex-name sweep
Last cycle an entry in this graph said nobody had checked whether the whole RPA complex - rather than its individual subunits - had ever been linked to the leukaemia gene KMT2A, and recorded that as an open question. We ran that check. It found a 2010 paper that measured RPA sitting on the exact stretch of the KMT2A gene where the cancer-causing breaks happen, after chemotherapy damage.
The same thing, for a scientist
**The open question, verbatim, from `interface-kmt2a-rpa2-declined-an-abundant-nuclear-ssdna-binder-on-six-spectra-but-the-resection-thread-is-live`:** *"Does 'RPA complex' as a search phrase, rather than the RPA2 symbol, return anything linking replication protein A to any MLL fusion? Rule 64 requires this and it was not run; the counted zero above is provisional until it is."* **Run 2026-09-24. Fifteen queries, every querytranslation read back, every non-zero list title-checked** [E1]. **Result: the subunit zeros hold, the complex-name query does not, and the difference is a real measurement.** `"replication protein A"[tiab] AND MLL[tiab]` returns **1** record - **PMID 21048951, Sun et al., PLoS One 2010** - and it is an **RPA ChIP-qPCR across the MLL/KMT2A breakpoint cluster region after etoposide**, with RPA enrichment transient in ATM-proficient cells and spatially extended plus persistent in ATM-deficient cells [E2]. `"RPA complex"[tiab]` against MLL and KMT2A returns 0; every subunit-symbol query against MLL returns 0 [E1]. **Two consequences.** (i) **The provisional status of the Q180 zero is discharged** - it was right at subunit level and rule 64's warning was justified, because the complex name found something the symbols could not. (ii) **Two absence entries in this corpus carry ids that over-generalise their own contents**: `absence-no-resection-factor-localised-at-the-kmt2a-locus` and `absence-no-damage-state-occupancy-measurement-for-any-resection-factor-at-the-kmt2a-locus` are both scoped, correctly and explicitly in their own `what_is_absent` fields, to six named proteins (BRCA1, BARD1, EXO1, DNA2, CBX3, KDM5A, plus CBX1). **RPA is not among them, and for RPA the damage-state locus measurement exists.** The fields are right; the ids and the headline framing are not [E3]. **This is reported, not edited - those entries belong to other beats.**
The fifteen queries, with counts and translations
Run with scripts/pubmed_safe.py so that no zero can be a silent instrument failure; every zero below carries a non-empty querytranslation and the 'No items found.' signature, and one tag per term (rule 94) 1.
| query | count | on topic |
|---|---|---|
"replication protein A"[tiab] AND MLL[tiab] | 1 | 1 - PMID 21048951 |
"replication protein A"[tiab] AND KMT2A[tiab] | 0 | - |
"RPA complex"[tiab] AND MLL[tiab] | 0 | - |
"RPA complex"[tiab] AND KMT2A[tiab] | 0 | - |
"replication protein A"[tiab] AND "MLL-AF4"[tiab] | 0 | - |
RPA1[tiab] AND MLL[tiab] | 0 | - |
RPA1[tiab] AND KMT2A[tiab] | 2 | 0 |
RPA2[tiab] AND MLL[tiab] | 0 | - |
RPA2[tiab] AND KMT2A[tiab] | 0 | - |
RPA3[tiab] AND MLL[tiab] | 0 | - |
RPA3[tiab] AND KMT2A[tiab] | 0 | - |
RPA32[tiab] AND MLL[tiab] | 0 | - |
RPA70[tiab] AND MLL[tiab] | 0 | - |
RFA2[tiab] AND MLL[tiab] | 0 | - |
"replication protein A"[tiab] AND leukemia[tiab] | 11 | 0 |
Positive controls, through the identical query shape (rule 67): "replication protein A"[tiab] alone = 1,569; MLL[tiab] AND KMT2A[tiab] = 296; RPA1[tiab] = 364, RPA2[tiab] = 324, RPA3[tiab] = 197. So the operands work and the zeros are about the conjunctions 1.
The 11-record leukemia list was opened and read, not counted - this project has been bitten by a control count nobody opened. All eleven titles are recorded in the data block; none concerns a fusion oncoprotein. Two are 1990s 'DNA synthesome' papers from leukaemia cells, one is the Curr Biol 2003 ATR/RPA translocation paper, one is a cohesin-mutant MDS/AML paper, and the rest are off-beat 1.
Why the symbol queries could not see it, which is the transferable lesson
Sun 2010 writes 'Replication protein A (RPA)' in its abstract and never uses a subunit symbol there. Its Methods name the antibody as 'anti-RPA34 (LabVision)' and its figure legends as 'anti-RPA32' - both are UniProt synonyms of P15927, so the protein detected is RPA2 2. The paper is therefore about RPA2 and is unreachable from RPA2[tiab].
That is rule 64 exactly - never write a counted absence on a complex SUBUNIT without running the complex form - and it is worth recording that the rule fired in the direction the rule predicts, on the beat that had written the rule into its own open questions. Q180 did the right thing: it flagged the query it had not run, in the entry, as provisional. The gap closed one cycle later because it was written down.
And the converse also held, which was not predicted. For the menin axis the asymmetry runs the other way: RPA2[tiab] AND menin[tiab] returns 6 records while "replication protein A"[tiab] AND menin[tiab] returns only 3, and RPA32[tiab] AND menin[tiab] returns 0 4. So the subunit symbol is the more sensitive probe for one axis and the complex name for the other. Rule 64 should be read as 'run both forms', not as 'the complex form is better'.
What the Sun 2010 result does and does not establish for this project
Does: there is a published, damage-state, locus-resolved measurement of RPA occupancy at the 8.3 kb MLL/KMT2A breakpoint cluster region spanning exons 7-13, around the ~20 bp therapy-related breakpoint hotspot, with a within-experiment control locus (beta-globin, 11p15), an ATM-deficient/proficient pair, a pharmacological ATM inhibitor arm, a bleomycin negative-agent arm, and three independent experiments for the RPA panels 2.
Does not:
- No fusion is present in any arm. The ATM comparison is a fibroblast pair (AT5BIVA and its chromosome-11-complemented derivative 11-4); the third line is BV173, a CML-derived B-cell line. This is about how the break is made, not about what the fusion protein does.
- Occupancy is not translocation. The authors propose the causal chain; they do not demonstrate it. This corpus's
gene-rad51already grades the same paper that way. - Their own control removes the obvious mechanism. BrdU incorporation after etoposide did not differ significantly between the BCR and the beta-globin control, so preferential replication-fork stalling does not explain the BCR-specific loading 2.
- It is a candidate-locus assay. Six primer sets and one control region; it cannot say the BCR is unusual relative to the genome, only relative to beta-globin.
- Single group, little follow-up in the sixteen years since.
And it was already in the corpus. PMID 21048951 is cited in gene-rad51, at evidence E2 and in a locus_level_measurement data block, whose own note reads 'the only measurement of any repair protein AT the KMT2A hotspot found in this beat'. So the paper was known and the absence entries were written anyway - because the absence entries enumerate proteins, and the paper is filed under a protein that is not on their list. That is search-the-citations-not-the-entry-text in its purest form: the refutation of a general claim was already cited in the same corpus, under another heading.
| script | no script - a query sweep with every count, querytranslation and idlist recorded inline, run through scripts/pubmed_safe.py so no zero can be a silent instrument failure |
|---|---|
| status | run |
| run date | 2026-09-24 |
| run by | Q192-the-rpa-heterotrimer |
| question answered | Open question 2 of interface-kmt2a-rpa2-declined-an-abundant-nuclear-ssdna-binder-on-six-spectra-but-the-resection-thread-is-live: 'Does RPA complex as a search phrase, rather than the RPA2 symbol, return anything linking replication protein A to any MLL fusion? Rule 64 requires this and it was not run; the counted zero above is provisional until it is.' ANSWERED: the subunit zeros hold at subunit level and are no longer provisional, and the complex-name form returns one record that the symbols could not reach - a damage-state RPA occupancy measurement at the KMT2A breakpoint cluster region (PMID 21048951). Note that the found paper is NOT an MLL-fusion interaction, which is what the open question asked about; the answer to the question as literally posed is still no. |
| result summary | 15 queries, 14 zeros, 1 on-topic hit. Complex-name form finds PMID 21048951 (RPA ChIP at the MLL 11q23 BCR after etoposide, ATM-dependent); no subunit-symbol form finds it. Two corpus absence entries carry ids broader than their scoped contents. The RPA2 decline's provisional zero is discharged and one of its four arguments (subunit coherence) is weakened by PMID 12509449. |
| absence class | measured-not-searched |
| absence class note | The residual RPA-versus-KMT2A absence is `measured-not-searched` and not `never-measured`, which is a change of class produced by this analysis: a damage-state RPA occupancy assay at this exact locus exists and has been performed once, so the question 'does RPA load at the KMT2A BCR in a KMT2A-REARRANGED blast' needs a repeat of a published protocol rather than a new method. Archive checks NOT run by this beat: GEO for an RPA or ssDNA occupancy dataset covering 11q23, and ProteomeXchange beyond PXD009338. Both are named as unrun rather than implied to be negative. |
inputs
[ "NCBI E-utilities esearch over PubMed via scripts/pubmed_safe.py", "PubMed efetch metadata for every non-zero idlist (titles read, not counted)", "PMC full text of PMID 21048951 (PMC2965082)", "Direct read of two absence entries and of gene-rad51 in this repository", "scripts/check-citation-reuse.py for PMIDs 21048951, 12509449, 42578372, 27625305, 25132265, 24267891, 12791985, 11927569, 23070815" ]
queries — the numbers behind the chart
[
{
"q": "\"replication protein A\"[tiab] AND MLL[tiab]",
"count": 1,
"translation": "\"replication protein A\"[Title/Abstract] AND \"MLL\"[Title/Abstract]",
"ids": [
"21048951"
],
"on_topic": 1
},
{
"q": "\"replication protein A\"[tiab] AND KMT2A[tiab]",
"count": 0,
"translation": "\"replication protein A\"[Title/Abstract] AND \"KMT2A\"[Title/Abstract]",
"genuine_zero": true
},
{
"q": "\"RPA complex\"[tiab] AND MLL[tiab]",
"count": 0,
"translation": "\"RPA complex\"[Title/Abstract] AND \"MLL\"[Title/Abstract]",
"genuine_zero": true
},
{
"q": "\"RPA complex\"[tiab] AND KMT2A[tiab]",
"count": 0,
"translation": "\"RPA complex\"[Title/Abstract] AND \"KMT2A\"[Title/Abstract]",
"genuine_zero": true
},
{
"q": "\"replication protein A\"[tiab] AND \"MLL-AF4\"[tiab]",
"count": 0,
"translation": "\"replication protein A\"[Title/Abstract] AND \"MLL-AF4\"[Title/Abstract]",
"genuine_zero": true
},
{
"q": "RPA1[tiab] AND MLL[tiab]",
"count": 0,
"translation": "\"RPA1\"[Title/Abstract] AND \"MLL\"[Title/Abstract]",
"genuine_zero": true
},
{
"q": "RPA1[tiab] AND KMT2A[tiab]",
"count": 2,
"translation": "\"RPA1\"[Title/Abstract] AND \"KMT2A\"[Title/Abstract]",
"ids": [
"41083063",
"35117846"
],
"on_topic": 0
},
{
"q": "RPA3[tiab] AND MLL[tiab]",
"count": 0,
"translation": "\"RPA3\"[Title/Abstract] AND \"MLL\"[Title/Abstract]",
"genuine_zero": true
},
{
"q": "RPA3[tiab] AND KMT2A[tiab]",
"count": 0,
"translation": "\"RPA3\"[Title/Abstract] AND \"KMT2A\"[Title/Abstract]",
"genuine_zero": true
},
{
"q": "RPA32[tiab] AND MLL[tiab]",
"count": 0,
"translation": "\"RPA32\"[Title/Abstract] AND \"MLL\"[Title/Abstract]",
"genuine_zero": true
},
{
"q": "RPA70[tiab] AND MLL[tiab]",
"count": 0,
"translation": "\"RPA70\"[Title/Abstract] AND \"MLL\"[Title/Abstract]",
"genuine_zero": true
},
{
"q": "RFA2[tiab] AND MLL[tiab]",
"count": 0,
"translation": "\"RFA2\"[Title/Abstract] AND \"MLL\"[Title/Abstract]",
"genuine_zero": true
},
{
"q": "\"replication protein A\"[tiab] AND leukemia[tiab]",
"count": 11,
"translation": "\"replication protein A\"[Title/Abstract] AND \"leukemia\"[Title/Abstract]",
"ids": [
"20301779",
"33673424",
"33351783",
"27869555",
"12814551",
"11173860",
"10814674",
"9882301",
"9774387",
"9419424",
"9279361"
],
"on_topic": 0
},
{
"q": "\"replication protein A\"[tiab]",
"count": 1569,
"translation": "\"replication protein A\"[Title/Abstract]",
"role": "positive control / denominator"
},
{
"q": "MLL[tiab] AND KMT2A[tiab]",
"count": 296,
"translation": "\"MLL\"[Title/Abstract] AND \"KMT2A\"[Title/Abstract]",
"role": "positive control on the fusion-side operand"
}
]the eleven leukemia titles read not counted
[ "20301779 - Dyskeratosis Congenita and Related Telomere Biology Disorders (GeneReviews, Seattle, 1993-; title obtained by NCBI esummary because the PubMed MCP metadata tool returned NO RECORD for this PMID - a book chapter, not a journal article, and an instrument limitation worth noting)", "33673424 - A Role for Human DNA Polymerase lambda in Alternative Lengthening of Telomeres (Int J Mol Sci 2021)", "33351783 - Cohesin mutations alter DNA damage repair and chromatin structure and create therapeutic vulnerabilities in MDS/AML (JCI Insight 2021)", "27869555 - The Influence of the CTIP Polymorphism Q418P on Homologous Recombination and Predisposition to Radiation-Induced Tumorigenesis in Mice (Radiat Res 2016)", "12814551 - ATR kinase activity regulates the intranuclear translocation of ATR and RPA following ionizing radiation (Curr Biol 2003) - ON TOPIC FOR RPA LOCALISATION, not for KMT2A; it is one of UniProt's ECO:0000269 citations for RPA2's nuclear and PML-body annotation", "11173860 - BLM, the Bloom's syndrome protein, varies during the cell cycle (Cytogenet Cell Genet 2000)", "10814674 - Cancer risk and the ATM gene: a continuing debate (JNCI 2000)", "9882301 - Human papillomavirus DNA replication compartments in a transient DNA replication system (J Virol 1999)", "9774387 - An eukaryotic RuvB-like protein (RUVBL1) essential for growth (JBC 1998)", "9419424 - The biochemical status of the DNA synthesome can distinguish between permanent and temporary cell growth arrest (Cell Growth Differ 1997)", "9279361 - The isolation of a DNA synthesome from human leukemia cells (Leuk Res 1997)" ]
corrections reported not applied
[
{
"entry": "absence-no-resection-factor-localised-at-the-kmt2a-locus",
"owner_beat": "q32-brca1-hr-licensing",
"what_is_wrong": "the id and the title generalise to 'no resection factor', while data.what_is_absent correctly scopes the claim to BRCA1, BARD1, EXO1, DNA2, CBX3/HP1-gamma and KDM5A",
"what_it_should_say": "the scoped form, e.g. 'no BRCA1, BARD1, EXO1, DNA2, HP1-gamma or KDM5A has been localised at the KMT2A locus' - which is what the title's own subtitle already says. The fields need no change.",
"citation_that_settles_it": "PMID 21048951 / 10.1371/journal.pone.0013554, Sun et al., PLoS One 2010 - RPA and RAD51 ChIP at the MLL 11q23 breakpoint cluster region after etoposide",
"read_or_abstract_only": "FULL TEXT READ from PMC2965082 on 2026-09-24",
"severity": "low - the scoped claim is intact; the risk is that the id gets quoted"
},
{
"entry": "absence-no-damage-state-occupancy-measurement-for-any-resection-factor-at-the-kmt2a-locus",
"owner_beat": "q35-encode-reanalysis",
"what_is_wrong": "the id says 'for ANY resection factor' and the title says the damage state is 'never-measured for all six proteins in every cell type on Earth'; the six are named and RPA is not one of them, but a damage-state occupancy measurement at this locus DOES exist for RPA (and for RAD51 and INO80). The entry's own open question - 'Is there any deposited dataset, anywhere, in which a resection factor was ChIPped in a cell after DNA damage at a mapped locus?' - is answered in the affirmative by a published ChIP-qPCR, though not by a DEPOSITED dataset.",
"what_it_should_say": "scope the id and headline to the named proteins, and record PMID 21048951 as the existing damage-state measurement at this locus for RPA/RAD51/INO80, which also supplies the assay design (six primer sets across the BCR, beta-globin control at 11p15) that a future damage-state ChIP in a KMT2A-rearranged cell should reuse",
"citation_that_settles_it": "PMID 21048951 / 10.1371/journal.pone.0013554",
"read_or_abstract_only": "FULL TEXT READ from PMC2965082 on 2026-09-24",
"severity": "medium - this entry exists specifically to stop a prior absence being re-quoted in a stale form, so an over-general id in it is the most likely to propagate"
},
{
"entry": "interface-kmt2a-rpa2-declined-an-abundant-nuclear-ssdna-binder-on-six-spectra-but-the-resection-thread-is-live",
"owner_beat": "q180-the-af4-specific-interactors",
"what_is_wrong": "two things. (1) Evidence E5 states 'RPA2 has no independent literature linking it to KMT2A' and that 'RPA2's 13 appearances in this corpus are all as a resection or HR assay readout, never as a KMT2A partner'. The first half is correct. The second is not: one of RPA2's corpus appearances is gene-rela evidence E5, a MENIN-RPA2 INTERACTION with functional data (PMID 28007956) - menin being the central node of this project - and that is not a resection readout. (2) Evidence E3's subunit-coherence argument states that a genuine RPA-complex association should have brought RPA1; the founding menin-RPA2 paper reports menin binding PREFERENTIALLY TO FREE RPA2 rather than the heterotrimer or an RPA2-RPA3 subcomplex, so for a menin-bridged association single-subunit recovery is the predicted outcome.",
"what_it_should_say": "the decline's VERDICT should stand - 6 spectra in two replicates is low and the nuclease and abundance arguments are untouched - but E5's corpus-grep claim should be narrowed, and E3 should note that the obligate-heterotrimer premise is qualified by the free-RPA2 preference in PMID 12509449 and by the existence of aRPA (RPA1/RPA3/RPA4)",
"citation_that_settles_it": "PMID 12509449 / 10.1128/MCB.23.2.493-509.2003, Sukhodolets et al., Mol Cell Biol 2003; and PMID 28007956 / 10.1093/carcin/bgw123, Chen et al., Carcinogenesis 2017, already cited in this corpus at gene-rela E5",
"read_or_abstract_only": "PMID 12509449 full text read from PMC151531; PMID 28007956 abstract only",
"severity": "medium - the verdict is unchanged, the stated reasons are partly wrong, and the entry is the corpus's entry of record for RPA2"
}
]measurements — the numbers behind the chart
[
{
"quantity": "count",
"subject": "on-topic records linking replication protein A to MLL or KMT2A in title/abstract",
"value": 1,
"units": "records",
"n": 15,
"universe": "15 PubMed queries over RPA complex names and subunit symbols against MLL, KMT2A and MLL-AF4, every querytranslation read back, 2026-09-24",
"null": 0,
"note": "null is the count returned by all thirteen other conjunctions; the denominators are 1,569 for the RPA operand and 296 for MLL AND KMT2A, so the operands work"
},
{
"quantity": "count",
"subject": "query forms that can reach PMID 21048951",
"value": 1,
"units": "of 12 forms tried against MLL or KMT2A",
"n": 12,
"universe": "the twelve MLL/KMT2A conjunctions in this sweep",
"null": null,
"note": "only the complex-name form reaches it; the paper's abstract writes 'Replication protein A (RPA)' and no subunit symbol, and its Methods use the RPA34 spelling which was not in this beat's name list"
},
{
"quantity": "count",
"subject": "corpus entries whose id over-generalises their own scoped absence, found by this beat",
"value": 2,
"units": "entries",
"n": 3086,
"universe": "the two resection absence entries surfaced by scripts/beat-inventory.py on keywords rpa/resection/ssdna and then read in full, out of 3,086 entries",
"null": null,
"note": "this is a found-2-while-looking-at-3 figure, not a corpus audit, and must not be quoted as a rate. No systematic scan for over-general absence ids was run."
}
]- Sun 2010's therapeutic implication runs OPPOSITE to the usual DNA-repair argument in oncology, and this is the bite-back that matters: their model is that excessive RPA and RAD51 loading at the KMT2A hotspot after etoposide CAUSES the therapy-related translocation, so inhibiting HR might reduce secondary leukaemia rather than sensitise the tumour. Any proposal to combine an HR-directed agent with etoposide in this disease inherits both directions of that argument at once.
- If RPA loading at the BCR is the licensing step for the translocation, then anything that raises replication stress or lowers the RPA reserve during etoposide exposure is a candidate risk factor for therapy-related KMT2A-rearranged leukaemia. That includes ATR inhibitors, which act by exhausting RPA.
- Does RPA load at the KMT2A breakpoint cluster region in a KMT2A-REARRANGED blast, and does the fusion protein's presence on the retained allele change it? Sun 2010 supplies the whole protocol - six primer sets across the BCR, beta-globin at 11p15 as control, etoposide 100 uM for 10 min, ChIP at 0 and 30 min - and no arm of it had a fusion. This is the cheapest high-value experiment this beat identifies.
- Is there a deposited genome-wide ssDNA or RPA occupancy dataset covering chr11q23? Not searched here. GEO and the END-seq/BLESS/S1-END-seq literatures were not queried, and this corpus's own absence entries already flag that gap for other proteins.
- The 'anti-RPA34 (LabVision)' antibody in Sun 2010 has no clone recorded. RPA34 is a UniProt synonym of RPA2 (P15927) AND a curated HGNC alias of POLR1G, so the reagent name alone is ambiguous between the ssDNA clamp and an RNA-polymerase-I subunit. The figure legends say RPA32, which resolves it - but is the clone recoverable, so that a repeat uses the same reagent?
- Was the systematic question ever asked: how many of this corpus's 79 absence- entries carry an id broader than the claim in their own fields? Two were found here while reading three, which is not a rate, and a scan has never been run.
- E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
Fifteen PubMed queries pairing RPA complex-level names and subunit symbols against MLL, KMT2A and MLL-AF4 were run on 2026-09-24 through scripts/pubmed_safe.py with one [tiab] tag per term and every querytranslation read back; the only non-zero on-topic result is one record for "replication protein A"[tiab] AND MLL[tiab] (PMID 21048951), while "RPA complex" against both MLL and KMT2A returns zero and every subunit symbol against MLL returns zero; positive controls through the identical shape return 1,569 for "replication protein A"[tiab] and 296 for MLL[tiab] AND KMT2A[tiab].
What this evidence cannot show
Bounded by TITLES AND ABSTRACTS (rule 80). A paper that measured RPA in a KMT2A-rearranged cell but mentioned neither RPA nor MLL/KMT2A in its abstract is invisible to all fifteen queries - and for a repair protein used as a readout, a supplementary figure is exactly where such a measurement lives, so the bound is weaker than the count suggests. Name list run on the RPA side: 'replication protein A', 'RPA complex', RPA1, RPA2, RPA3, RPA32, RPA70, RFA2. NOT run: RPA34, RPA14, REPA1, REPA2, REPA3, HSSB, 'RP-A', 'RF-A', 'single-stranded DNA binding protein'. Fusion side: MLL, KMT2A, 'MLL-AF4'. NOT run: the other partner fusions, or 't(4;11)' (which would need quoting per rule 95). One archive only - PubMed; no GEO, PRIDE or ProteomeXchange query is part of this item.
Evidence that argues against this (1)
- Fifteen queries returning fourteen zeros looks like a thorough absence and is partly an artefact of query form: every one of these conjunctions requires two specific strings in one abstract, and the single hit demonstrates that the literature's actual habit is to name the complex and not the subunit. A name list that missed 'RPA34' - the very spelling the one positive paper uses in its Methods - would have missed the paper had the complex form not been run. The honest reading is that this sweep is sensitive enough to have found one paper and is not sensitive enough to bound the field. This repository 2026
- E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
Sun et al. 2010 performed chromatin immunoprecipitation with anti-RPA antibodies across the MLL breakpoint cluster region after etoposide treatment: ATM-proficient 11-4 cells showed etoposide-induced RPA enrichment in the BCR immediately after etoposide removal which returned to normal by 30 minutes, while ATM-deficient AT5BIVA cells showed enrichment that was more spatially extensive (extending into regions adjacent to bt56/t56) and that remained high at 30 minutes; RPA expression was not altered by etoposide or by ATM status; RAD51 and INO80 were also enriched at the same region and more so in ATM-deficient cells; bleomycin, which is not associated with 11q23 translocations, did not produce a significant increase; and BrdU incorporation after etoposide did not differ significantly between the BCR and the beta-globin control locus.
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, including the supplementary figure legends; the figures themselves were not viewed and no effect size was extracted, because the PMC text renders the figure panels as empty cross-references - so every quantitative statement here is the authors' prose, not a number read off a plot. NO FUSION IN ANY ARM: AT5BIVA is an SV40-transformed ATM-deficient human fibroblast line, 11-4 is its chromosome-11-complemented derivative, BV173 is a CML-derived B-cell line. Etoposide 100 uM for 10 min is a pharmacological dose. Candidate-locus ChIP-qPCR, six primer sets, one control region (beta-globin, 11p15) - no genome-wide comparison. Antibody named 'anti-RPA34 (LabVision)' in Methods and 'anti-RPA32' in figure legends, both P15927 synonyms, no clone given (rule 74); the paper's Figure S13 shows it immunoprecipitates both phosphorylated and unphosphorylated RPA32. Three independent experiments for the RPA panels. Single group; little follow-up since 2010.
Evidence that argues against this (1)
- The paper's own BrdU control removes the mechanism a reader will assume, and that is the most important limitation to carry: fork stalling at the BCR is NOT elevated relative to the control locus, so 'the BCR is a replication-fragile site and therefore RPA accumulates there' is not what the data show. The authors say the nature of the etoposide-induced damage in the BCR is still unclear. Occupancy is also not translocation - the causal step from 'RPA and RAD51 over-load here' to 'a translocation forms' is a proposal. Sun J et al. 2010
- E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database
Two absence entries in this corpus carry ids and headline framings broader than their own scoped contents: absence-no-resection-factor-localised-at-the-kmt2a-locus and absence-no-damage-state-occupancy-measurement-for-any-resection-factor-at-the-kmt2a-locus both restrict themselves, in their own data.what_is_absent and absence_class_per_protein fields, to BRCA1, BARD1, EXO1, DNA2, CBX3/HP1-gamma, KDM5A and (added by Q35) CBX1 - and RPA is not among them - while their ids say 'no resection factor' and 'any resection factor'; for RPA, a damage-state occupancy measurement at the KMT2A locus is published (PMID 21048951) and is already cited in this corpus at gene-rad51 evidence E2.
What this evidence cannot show
THE SCOPED CLAIMS ARE NOT REFUTED and this item must not be read as saying they are. Both entries are explicit about their six or seven proteins, and for those proteins the damage-state absence stands: nothing in Sun 2010 measures BRCA1, BARD1, EXO1, DNA2, CBX3, CBX1 or KDM5A. What is wrong is the GENERALITY of the id and of the title, and ids are load-bearing in this graph because they are what gets quoted and linked. Reported here, not edited - those entries belong to Q32 and Q35 and the standing rule is to report rather than edit another beat's files. A second, weaker point: Sun 2010 is ChIP-qPCR at a candidate locus and not a deposited genome-wide dataset, so if the intended claim in those entries was specifically about ARCHIVES rather than about the literature, the id is still over-general but the archive census inside them is untouched.
Evidence that argues against this (1)
- The counter-case for my own correction: 'resection factor' may legitimately exclude RPA in the authors' vocabulary, because RPA is a ssDNA-binding protein that ACTS AFTER resection rather than an exonuclease or nuclease-licensing factor - so a specialist might read 'no resection factor' as a claim about the nuclease layer and consider RPA out of scope by definition. That reading is defensible. It does not rescue the id, because a reader arriving from the graph cannot know which vocabulary is intended, and this corpus's own resection thread names RPA2 thirteen times as the READOUT of resection, which makes RPA part of the topic by the corpus's own usage. This repository 2026
- E4established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database
The complex-name and subunit-symbol query forms are asymmetric in OPPOSITE directions for the two RPA axes relevant to this project: for the KMT2A-locus axis the complex name finds one record and every subunit symbol finds none, while for the menin axis RPA2[tiab] AND menin[tiab] returns 6 records against 3 for "replication protein A"[tiab] AND menin[tiab] and 0 for RPA32[tiab] AND menin[tiab].
What this evidence cannot show
Two axes and three query forms is a small basis for a claim about query behaviour; this is an observation from one beat, not a measured rule. The mechanism is nonetheless legible: the menin literature is a protein-interaction literature that necessarily names which subunit binds, while the locus literature is a chromatin literature that names the complex it immunoprecipitated. So the direction of the asymmetry is predictable from the KIND of experiment, which is a more useful generalisation than the counts.
Evidence that argues against this (1)
- The six-record menin count is inflated as evidence: two of the six merely LIST RPA2 among menin's partners (a MEN1 review and a MEN1 expression-profiling paper), one is pure computation, and one is the 2026 menin-HR paper that names 'the RPA2/Rad51 pathway' as a downstream step rather than reporting a menin-RPA2 contact. Stripped to independent experimental work on the contact itself, the count is 2, not 6 - so the asymmetry in SENSITIVITY is real and the asymmetry in EVIDENCE is much smaller. Agarwal SK et al. 2004
runs the rule-64 complex-form query that entry flagged as unrun; the subunit zeros hold and the complex name finds PMID 21048951
the id and headline generalise to 'any resection factor'; a damage-state RPA/RAD51/INO80 occupancy measurement at this locus is published. The SCOPED six-protein claim inside the entry is untouched.
same over-generality in the id; the six named proteins remain unlocalised at the locus
supplies the locus-level measurement recorded in the complex entry's data block