Oncofusion

start / graph / ana-q120-the-mll1-n-channel-is-inert-to-three-decimals-in-both-human-lines

Analysisstrong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.Q120-h3k79me2-second-line-and-the-mll1-epitope

At the MLL1-N HELD SET, floor-normalised, the MLL1-N channel is inert to three decimals in both human lines (1.036 and 1.0359) - AND THAT IS A STATEMENT ABOUT THAT POPULATION, NOT ABOUT THE CHANNEL: at MLL1's own top sites in mouse the same channel reads 0.3012 under a DOT1L inhibitor and 0.3298 under a menin inhibitor

also called 1.036 twice · the MLL1-N channel is inert in both lines

The antibody most papers use to ask whether the fusion protein is still on DNA gives the same non-answer in two different leukaemia cell lines, to three decimal places. When a measurement is that reproducible and that flat, the thing being reproduced is usually the tool.

The same thing, for a scientist

At each line's own MLL1-N held set (top 0.1% of 2 kb bins by MLL1-N DMSO signal), floor-normalised against that track's own aggregate share ratio over ~500,000 mid-occupancy bins [E1]: | | RS4;11 (KMT2A::AFF1, B-ALL) | MOLM13 (KMT2A::MLLT3, AML) | |---|---:|---:| | MLL1-N share ratio | 0.9069 | 0.9438 | | its own genome-wide floor | 0.8753 | 0.9111 | | **floor-normalised** | **1.036** | **1.0359** | | per-bin median share ratio | - | 0.9991 | | fraction still in its own top 0.1% | 0.4367 | 0.6645 | **Two cell lines, two fusion partners, two lineages, n = 1 each, and the floor-normalised answer is the same to the fourth significant figure.** The agreement is almost certainly partly luck at n = 1; what is not luck is that both are indistinguishable from 1.000, i.e. from no change at all, while menin at the same bins in the same cells is at 0.256 and 0.258. **The reading this forces.** Bethyl A300-086A, deposited as 'anti-MLL1n', an MLL1 N-TERMINUS epitope. The N-terminus is retained in every KMT2A fusion, so this track is the sum of the fusion and wild-type MLL1 in unknown proportion and CANNOT distinguish them. The immunogen residue range was not verified: three vendor routes were tried this session (fortislife.com two URL forms, bethyl.com/product/A300-086A) and all returned HTTP 404, so 'N-terminus' rests on the depositors' own target label and on the standard fact that MLL-FPs lack the C-terminal SET domain. A channel that is the sum of a menin-dependent species and a menin-independent one will move by the fusion's share of the sum. Measuring 1.036 twice, in two lines whose fusion burden and wild-type MLL1 burden have no reason to be identical, is more consistent with the fusion contributing a small fraction of the signal in both than with the fusion genuinely staying put in both. **It is also the strongest available argument for the experiment this graph keeps costing and nobody has run:** a tagged fusion in a human KMT2A-rearranged line. Q118 wrote it for RS4;11; this entry says the same experiment in MOLM13 would now discriminate two models that the MLL1-N channel cannot, because the mouse arm of this beat shows what a tag does to the same question.

Go deeper
An independent re-derivation of a sibling beat's number

Q118 recorded MOLM13 MLL1-N held-set share ratio 0.9438, per-bin median 0.9991 and fraction still in own top 0.6645. This beat re-downloaded GSM3635685 and GSM3635686, re-binned them from scratch and got 0.9438, 0.9991 and 0.6645.

That is worth recording on its own. The project's standing worry is stale or unre-derivable numbers; here a number from the previous cycle was re-derived end to end from the deposited files and came back identical, which is the cheapest possible evidence that the pipeline is deterministic and the stored figures are not drifting.

QUALIFIED BY THIS BEAT'S OWN STRATIFICATION - the inertness is an aggregate

ana-q120-every-channel-moves-more-at-the-strongest-bins-and-the-mll1-n-inertness-is-an-aggregate splits the MOLM13 sets by signal amplitude and finds the MLL1-N channel is not uniformly inert: at the strongest tenth of the MLL1-N held set the share ratio is 0.6919, and at the strongest tenth of the co-occupied set 0.5304, against aggregates of 0.9438 and 0.8020. The flat aggregate is produced by the weaker two thirds of each set.

So read the 1.036 figures as aggregate statements. The reagent argument they support is unchanged in kind - a channel that is the sum of a menin-dependent and a menin-independent species will move by the fusion's share of the sum, and that share plainly varies across sites - but the words 'does not move' apply to the set as a whole and not to its strongest members. The equivalent stratification for RS4;11 has not been computed by anyone.

Facts and figures
scriptanalyses/q120-h3k79me2-second-line-and-the-mll1-epitope/arm1-molm13-ordering.py
statusrun
result summaryMLL1-N floor-normalised share ratio is 1.036 in RS4;11 and 1.0359 in MOLM13, i.e. no change in either line, while menin at the same bins is 0.256/0.258.
antibodyBethyl A300-086A, deposited as anti-MLL1n, lots 5 or 6
epigenetic readoutthe MLL1-N ChIP channel is flat under a menin inhibitor in both human lines; H3K79me2 at the same bins is 0.7008 in RS4;11 and 0.9980 in MOLM13
shown for fusionKMT2A::AFF1 in RS4;11 and KMT2A::MLLT3 in MOLM13 - and in neither case can the reagent attribute the flat signal to the fusion rather than to wild-type MLL1
absence classnever-measured
absence class noteThe gap this entry names is NOT closeable from any archive: no ChIP of a TAGGED KMT2A fusion in any human KMT2A-rearranged line exists in GEO, so the fusion-versus-wild-type split of this channel has never been measured in human cells by anyone. It needs a knock-in or a tagged allele, i.e. wet work. The murine GSE260742 FLAG arm is the nearest available substitute and is a different system.
absence class assigned2026-09-23 (Q120)
TITLE QUALIFIED CYCLE 84
{
  "what_the_title_said": "The MLL1 N-terminus channel does not move under VTP-50469 in either human line, and the two floor-normalised figures agree to three decimals - 1.036 in RS4;11 and 1.0359 in MOLM13 - which makes the reagent's inertness a property of the reagent and not of the fusion",
  "why_it_changed": "the measurement in this entry is sound and is unchanged. Its TITLE asserted inertness of a channel without naming the population, and a title is the part that travels - a failure mode this corpus has recorded twice (a warning in a field does not protect the prose; the last line is the one that gets quoted).",
  "the_second_measurement": "Q122's criterion C7, curated in cycle 83: murine MLL-AF9 F3_NT at S_MLL1 reads 0.3012 under 5 uM SGC0946 and 0.3298 under 500 nM VTP-50469, against a pre-registered floor of 0.85.",
  "NOT_A_REFUTATION": "the two quantities differ in THREE ways at once - floor-normalised versus raw within-sample share ratio, MLL1-N held set versus the channel's own top sites, human versus mouse. Any one could account for the gap, and the reconciling arithmetic is on bins already on disk and has not been run.",
  "the_id_was_NOT_changed": "renaming the id would touch every edge pointing at it. The id still contains the word 'inert'; the title no longer asserts it unqualified.",
  "see": "ana-q122-curated-the-mll1-channel-is-not-inert-at-its-own-sites-and-a-dot1l-inhibitor-moves-it-as-much-as-a-menin-inhibitor"
}
How this could bite us back
  • If the MLL1-N channel is mostly wild-type MLL1 in both lines, then the field's standard occupancy assay for menin-inhibitor pharmacodynamics is measuring the protein the drug is not aimed at, and every 'the fusion is retained' statement built on it is unsupported rather than wrong.
  • The subtraction experiment above is cheap and nobody has done it, which is itself a warning: a field converges on one reagent and the control that would qualify it stops being run.
Still open
  1. What fraction of the MLL1-N ChIP signal in a human KMT2A-rearranged line is the fusion? The murine FLAG arm gives an upper bound of 0.73 in an over-expression model; the human number is unmeasured and every human occupancy statement in this literature depends on it.
  2. Would an MLL1 C-TERMINAL antibody - an epitope LOST in the fusion - settle it by subtraction? An N-terminal and a C-terminal ChIP in the same cells give fusion-plus-wild-type and wild-type, and their difference is the fusion. No sample in GSE127507 or GSE260742 uses a C-terminal MLL1 antibody.
Every source behind this entry (1)
  1. E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.dataset

    MLL1-N held-set share ratios under 330 nM VTP-50469 are 0.9069 in RS4;11 (floor 0.8753, normalised 1.036) and 0.9438 in MOLM13 (floor 0.9111, normalised 1.0359), with MOLM13 per-bin median 0.9991 and 0.6645 of held bins still in the track's own top 0.1%; menin at the same MLL1-N held bins is 0.2560 and 0.2578 respectively.

    What this evidence cannot show

    The RS4;11 figures are Q118's, re-read from its committed function-arm.json. NO SPIKE-IN NORMALISATION. All 26 GSE127507 records state a Drosophila spike-in was run, measured, and deliberately not used, and no scale factor is deposited; no GSE260742 record contains spike-in vocabulary at all. Every cross-condition number here is therefore a WITHIN-SAMPLE share or rank, which cancels the unknown per-sample scale factor exactly. No raw fold change is reported as a biological quantity anywhere in this beat. n = 1 per antibody per condition per line in both series. No number in this beat carries a spread and none is precise. Bethyl A300-086A, deposited as 'anti-MLL1n', an MLL1 N-TERMINUS epitope. The N-terminus is retained in every KMT2A fusion, so this track is the sum of the fusion and wild-type MLL1 in unknown proportion and CANNOT distinguish them. The immunogen residue range was not verified: three vendor routes were tried this session (fortislife.com two URL forms, bethyl.com/product/A300-086A) and all returned HTTP 404, so 'N-terminus' rests on the depositors' own target label and on the standard fact that MLL-FPs lack the C-terminal SET domain. A floor-normalised ratio of 1.036 in both lines is an agreement between two single measurements and the fourth-figure precision should not be read as a precision claim.

    Evidence that argues against this (1)
Connected to