start / graph / ana-q143-in-the-revumenib-exposed-relapse-sample-no-channel-is-enriched-above-a-non-specific-reference

AnalysismoderateQ143-menin-under-drug-in-infants

In the revumenib-exposed relapse sample nothing is enriched above a non-specific reference - not menin, not the fusion N-terminus, not ENL - and the sample's own top-ranked promoters are mitochondrial pseudogenes

In the cells taken while the child was on the drug, none of the five proteins measured is sitting anywhere in particular. The genes that score highest are the ones that score highest in a negative control. So the question 'is menin still where the fusion is' has no answer here, because the fusion is not measurably anywhere either.

The same thing, for a scientist

Ranking 152985's own promoters on its own KMT2A-N signal and dividing each channel's enrichment by an out-of-sample IgG reference gives, at the top 50: **menin 0.42x, KMT2A-N 0.33x, KMT2A-C 0.31x, ENL 0.80x, DOT1L 0.98x** - every channel **at or below** the non-specific reference, and the ordering stable across k = 50, 200, 1000 and 5000. The sample's top-ranked promoters are `MTRNR2L8, MTRNR2L2, TSEN34, FARP2, TMEM87B, FAM86B2, MBOAT7, FBRS, G3BP1, ZBTB20` - nuclear-mitochondrial pseudogenes and housekeeping loci, the signature of a library with no specific signal - and **0 of the top 50 are on the authors' oncoprotein-target list**, against 40 of 50 in the drug-free aliquot of the same patient. The three CUT&Tag channels of this sample also fail their own replicate-agreement check, so a library-quality explanation and a target-absent explanation both fit and this deposit separates neither.

Go deeper

The table, both samples side by side

Accessibility-corrected enrichment at each sample's own top-k KMT2A-N promoters 1:

	k=50	k=200	k=1000	k=5000
148752 menin	2.73	1.64	1.26	1.09
148752 KMT2A-N	3.12	1.64	1.32	1.19
148752 ENL	11.76	2.76	1.70	1.36
152985 menin	0.42	0.38	0.39	0.46
152985 KMT2A-N	0.33	0.35	0.39	0.51
152985 KMT2A-C	0.31	0.34	0.39	0.51
152985 ENL	0.80	0.82	0.85	0.84
152985 DOT1L	0.98	0.76	0.72	0.71

ENL and DOT1L are the least depleted in the drug arm, and neither reaches 1.0.

Why 'at or below IgG' rather than 'reduced by X%'

A below-1.0 value does not mean the protein is less present than a non-specific antibody. It means that after each channel is normalised to its own library total, the channel's enrichment gradient across promoters is flatter than the IgG reference's. With no in-sample IgG, no spike-in, and a different-clone comparator, the only honest statement is a qualitative one: nothing in this sample identifies a specific promoter set. See the confound entry and the replicate-agreement entry for why a quantitative statement is not available.

What the 0 of 50 does NOT establish

It does not establish that menin or the fusion have left chromatin. Three readings survive: the target is gone; the target is present but at a level below what this library can resolve; or the library is poor. The MLL::ENL transcript is still expressed in this sample by the authors' RT-PCR, which makes the middle reading live. A within-library share cannot measure an absolute amount, so 'undetectable in a normalised track' and 'absent from chromatin' are different claims and only the first is supported.

Facts and figures

sample	152985, relapse infant AML on revumenib, KMT2A::MLLT1
absence class	measured-not-analysed
status	run
script	analyses/menin-under-revumenib-in-an-infant-patient-pair/run2_controls.py

accessibility corrected enrichment at own topk

oncoprotein target hits in top50

accessibility corrected enrichment at own topk — the numbers behind the chart

{
  "k50": {
    "MENIN": 0.42,
    "KMT2A_N": 0.33,
    "KMT2A_C": 0.31,
    "ENL": 0.8,
    "DOT1L": 0.98
  },
  "k200": {
    "MENIN": 0.38,
    "KMT2A_N": 0.35,
    "KMT2A_C": 0.34,
    "ENL": 0.82,
    "DOT1L": 0.76
  },
  "k1000": {
    "MENIN": 0.39,
    "KMT2A_N": 0.39,
    "KMT2A_C": 0.39,
    "ENL": 0.85,
    "DOT1L": 0.72
  },
  "k5000": {
    "MENIN": 0.46,
    "KMT2A_N": 0.51,
    "KMT2A_C": 0.51,
    "ENL": 0.84,
    "DOT1L": 0.71
  }
}

oncoprotein target hits in top50 — the numbers behind the chart

{
  "152985_KMT2A_N": 0,
  "152985_MENIN": 0,
  "148752_KMT2A_N": 40,
  "expected_by_chance": 1.77
}

top promoters

[
  "MTRNR2L8",
  "MTRNR2L2",
  "TSEN34",
  "FARP2",
  "TMEM87B",
  "FAM86B2",
  "MBOAT7",
  "FBRS",
  "G3BP1",
  "ZBTB20"
]

three surviving readings

[
  "the targets have left chromatin",
  "the targets are present below this library's resolution",
  "the libraries are poor"
]

How this could bite us back

'Menin is undetectable under a menin inhibitor in a patient' is the sentence this dataset will be cited for. It is a statement about a normalised track in a relapse clone with no spike-in and no IgG, and it is compatible with menin being present and merely redistributed.
If an occupancy readout is used as a clinical pharmacodynamic endpoint, note that in the one patient sample where it has been tried the assay returned background in every channel including the ones expected to be positive - so a negative result from such an assay cannot currently be distinguished from an assay failure.

Still open

Is the fusion protein present but displaced, or absent? The transcript is expressed; no protein measurement exists for this sample.
Would a spike-in have distinguished 'gone' from 'globally reduced'? None exists anywhere in this deposit, and that single omission is what makes every statement here relative.
Would down-sampling the drug-free libraries to the drug-arm depths reproduce the collapse? That control was not run and would partly separate library quality from biology.

Every source behind this entry (3)

E1moderate confidence 3 of 5: One strong study, or several smaller consistent ones.dataset
At every promoter set tested in the revumenib-exposed relapse sample, menin, KMT2A-N, KMT2A-C, ENL and DOT1L are all at or below an out-of-sample non-specific reference.
analyses/menin-under-revumenib-in-an-infant-patient-pair - pre-registered re-analysis of GSE252378 samples 148752 and 152985 · Q143-menin-under-drug-in-infants (Oncofusion Knowledge Graph) · this repository 2026
What this evidence cannot show
n = 1 aliquot. The reference is an IgG from a different patient with a different fusion, CUT&RUN only, so it is not a background control for the three CUT&Tag channels. Within-library shares only; no spike-in. The three CUT&Tag channels of this sample have replicate agreement of 0.11-0.25.
Evidence that argues against this (1)
- The paper's own quantification over the same peak intervals puts menin at 0.871 and ENL at 0.827 of their diagnostic-sample values - i.e. a mild reduction, not a collapse - so the two instruments give very different impressions of the same libraries, and the difference is the denominator. Janssens DH 2024
E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.dataset
0 of the top 50 promoters in this sample are on the authors' published oncoprotein-target list, against 40 of 50 in the drug-free aliquot of the same patient.
analyses/menin-under-revumenib-in-an-infant-patient-pair - pre-registered re-analysis of GSE252378 samples 148752 and 152985 · Q143-menin-under-drug-in-infants (Oncofusion Knowledge Graph) · this repository 2026
What this evidence cannot show
Hypergeometric against the 687 mapped target symbols in a 19,404-gene universe; a zero at k = 50 has a rule-of-three upper bound of about 6% true hit rate, against a 3.5% base rate, so it is consistent with chance and not with enrichment.
Evidence that argues against this (1)
- A zero at k = 50 is a small-sample statement: the rule of three puts the true hit rate below about 6%, against a base rate of 3.5%, so the data are consistent with a weak residual enrichment that this test cannot resolve, and 'zero of fifty' should not be read as 'nothing at all'. Q143-menin-under-drug-in-infants (Oncofusion Knowledge Graph) 2026
E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper
The paper reports that menin-inhibitor treatment reduced the genome-wide MLL N and C terminal signal to undetectable levels in this sample, and that menin and DOT1L were likewise reduced to undetectable levels.
MLL oncoprotein levels influence leukemia lineage identities · Janssens DH, Duran M, Otto DJ, Wu W, Xu Y, Kirkey D, Mullighan CG, Yi JS, Meshinchi S, Sarthy JF, Ahmad K, Henikoff S · Nature Communications 15:9341 2024
PMID 39472576 DOI
What this evidence cannot show
The authors' statement rests on genome-browser tracks and peak calls; 'undetectable' in a normalised track is not an absolute measurement, and the same authors' own quantitative table over their oncoprotein peak set gives menin 0.871 rather than zero.
Evidence that argues against this (1)
- The MLL::ENL oncogene is still expressed in this sample by the authors' own RT-PCR with two primer sets, so the fusion is being transcribed while its chromatin signal is not detectable - which makes a sensitivity explanation as available as an absence explanation. Janssens DH et al. 2024

Connected to

depends on →148752 and 152985 are one patient before and during revumenib, not two treated patients - so the drug-free arm this corpus said did not exist is sample 148752
the drug arm of the pair
depends on →The menin, ENL and DOT1L replicates of the drug-exposed relapse sample agree with each other at Spearman 0.11 to 0.25, below the floor this project set in the previous cycle, so no quantitative claim from that arm survives
the drug arm's replicate agreement bounds what can be claimed
depends on →Drug exposure and clonal replacement are completely confounded across this pair - the relapse sample is a different clone in a different lineage, and the paper's own reading is that the drug killed the cells the diagnostic sample was made of
the clonal-replacement confound
related to →In drug-naive primary infant KMT2A::MLLT1 B-ALL, menin is as enriched as the fusion N-terminus at the top fusion promoters (2.73x vs 3.12x over an accessibility reference) while the wild-type C-terminal reporter sits at background (1.10x)
the same measurement in the drug-free aliquot of the same patient
related to →Revumenib (SNDX-5613, Revuforj) — first approved menin–KMT2A interface inhibitor
the exposure the sample was drawn under
depends on →The compound in the only infant menin-inhibitor occupancy sample is revumenib, named in the abstract - and neither the dose nor the duration of exposure is recorded in the paper, the GEO record or the authors' repository
no dose or exposure is recorded for this sample
← depends onMenin does not outlast the fusion under revumenib in primary infant material - 0.586 against 0.485 at oncoprotein peaks, a ratio of 1.21 against a pre-registered threshold of 1.5 - so the fusion-independent menin plateau seen after a degron does not reproduce here
the drug arm is at background in every channel
← related toThe authors' own deposited table puts menin at 0.87 of its diagnostic-sample signal over their oncoprotein-target peaks, in the sample their text describes as having menin reduced to undetectable levels - the two statements are about different denominators, and only one of them travels
the same libraries read with a different denominator
← depends onIn a patient, the menin-fusion association is demonstrable before the drug and unmeasurable after it - and the asymmetry is what this pair establishes, not a rate of menin loss
the negative half
← depends onAn occupancy readout returned background in every channel of the only drug-exposed patient sample it has been tried in - so a negative from such an assay cannot currently be told apart from an assay failure
the measurement this risk is drawn from