Hypothesis: H3K79me2 at a KMT2A fusion's own sites survives menin inhibition when the C-terminal partner contacts DOT1L directly (MLLT3/AF9) and decays when it does not (AFF1) - so the mark's response is a property of the partner and not of menin, which predicts the opposite of the direct-contact intuition and is testable in one line

[
  "exp-q120-h3k79me2-chip-in-a-kmt2a-afdn-line-under-a-menin-inhibitor",
  "exp-q120-an-mll1-c-terminal-chip-beside-the-n-terminal-one-in-one-human-line"
]

[
  "ana-q120-three-of-the-four-channels-replicate-across-two-fusion-partners-and-h3k79me2-does-not",
  "ana-q120-the-sgc0946-arm-of-gse260742-would-bound-the-marks-decay-for-one-download"
]

[
  "PARTNER ARM, NOT YET RUN. In a KMT2A::AFDN (MLL-AF6) line - AFDN does NOT contact DOT1L directly, so the hypothesis predicts the RS4;11 behaviour - H3K79me2 at the MLL1-N-and-MENIN co-occupied 2 kb bins under a menin inhibitor at 3 days has an aggregate within-sample share ratio BELOW 0.80, floor-normalised. If it is at or above 0.90, i.e. MOLM13-like in a line that is AML but not AF9-family, the partner explanation is refuted and lineage wins. Indifference band 0.80-0.90; report the magnitude inside it rather than a verdict. REAGENT AND WHAT IT CANNOT DISTINGUISH: the mark must be read with an anti-H3K79me2 antibody (CST #5427 in GSE127507, Abcam ab3594 in GSE260742 - two DIFFERENT polyclonals, so the comparison must stay inside one series or carry the reagent switch as a caveat). Either reagent reads the MARK and cannot distinguish halted deposition from active demethylation from replicative dilution, and cannot attribute any of them to the fusion; and the co-occupancy set must be defined with Bethyl A300-086A, an MLL1 N-TERMINAL epitope RETAINED in the fusion, which cannot distinguish the fusion from wild-type MLL1. So this falsifier can tell partner from lineage and CANNOT tell whether the surviving mark is at fusion-bound or wild-type-bound chromatin.",
  "LINEAGE ARM, NOT YET RUN. In a myeloid line with NO KMT2A rearrangement - GSE260742 deposits K562 and OCIAML3 - H3K79me2 under a menin inhibitor at the top 0.1% of its own H3K79me2 domains has an aggregate share ratio ABOVE 0.90. If the mark is unmoved in fusion-free myeloid cells too, then MOLM13's persistence is a lineage property and this hypothesis's central contrast is explained without any partner. Band 0.85-0.95. REAGENT: as above, and additionally in a fusion-free cell there is no fusion to attribute anything to, which is the point of the control and also its limit - a null here does not say the fusion's sites behave the same way.",
  "MECHANISM ARM, NOT YET RUN AND THE CHEAPEST OF THE THREE. If the mark survives in MOLM13 because residual DOT1L is catalytically sufficient rather than because of a partner contact, then removing DOTIL CATALYSIS directly should collapse it. GSE260742 deposits an SGC0946 H3K79me2 arm in murine MLLAF9 (an AF9 fusion). Prediction: H3K79me2 at the FLAG-tagged fusion's own top 0.1% of bins under SGC0946 has a share ratio BELOW 0.50 while the same bins under VTP50469 stay above 0.85. If BOTH drugs leave the mark high, the mark is not DOT1L-limited on this timescale at all and both this hypothesis and the residual-enzyme reading are wrong. Band 0.45-0.55 and 0.80-0.90. REAGENT: Abcam ab3594 anti-H3K79me2 in that series - a mark, not the fusion - and the FLAG tag (NEB 14793S) which reads the tagged fusion only and is blind to wild-type MLL1 and to any fusion whose tag has been proteolysed."
]

{
  "the_problem": "LINEAGE TRACKS THE FUSION PARTNER IN EVERY SYSTEM THIS HYPOTHESIS HAS BEEN TESTED IN. RS4;11 (KMT2A::AFF1, retains the mark least at 0.6578) is B-LYMPHOBLASTIC. MOLM13 (KMT2A::MLLT3, 0.8667) and murine MLL-AF9 (KMT2A::MLLT3, 0.7734/0.8318) are MYELOID. So 'set by the fusion partner' and 'set by lineage' make identical predictions on every measurement that supports this hypothesis.",
  "what_the_evidence_does_establish": "that the difference is not noise. Cycle 83's curation of Q122's C3 shows the cross-partner delta (0.2089) is 3.2-3.6x the between-culture spread (0.0584/0.0655), C6 replicates MOLM13 in an independent series (0.9063 vs 0.8667), and C8 survives a top-decile stratification in 8 of 8 measurements. The EFFECT is real; its ATTRIBUTION is not settled.",
  "the_decisive_design_and_it_does_not_exist": "a partner-matched pair within one lineage - KMT2A::AFF1 and KMT2A::MLLT3 both in B-lineage, or both myeloid - or a lineage-matched pair with different partners. No deposit this project has found contains either.",
  "why_the_ID_WAS_NOT_RENAMED": "renaming touches every edge pointing at this hypothesis, and the hypothesis is not wrong - it is under-determined. Naming the confound inside the entry is the honest fix; a rename would imply the partner reading had been retired, which it has not.",
  "how_to_quote_this_hypothesis_from_now_on": "'whether H3K79me2 survives a menin inhibitor is set by the fusion partner OR BY THE LINEAGE THAT TRACKS IT, not by menin'. The 'not by menin' half is well supported; the partner-versus-lineage half is not decided by anything measured so far.",
  "recorded_by": "king, cycle 84, from cycle 83's curation of Q122 criteria C1/C2/C3/C6/C8"
}