Oncofusion

start / graph / exp-q148-spike-in-normalised-paired-n-and-c-cut-and-run-in-a-hoxa-negative-kmt2a-rearranged-blast

Experimentmoderate confidence 3 of 5: One strong study, or several smaller consistent ones.Q148-what-the-hoxa-negative-blast-occupies

Spike-in-normalised paired KMT2A N- and C-terminal CUT&RUN, with a fusion-junction readout, in a HOXA-negative KMT2A-rearranged blast beside a HOXA-positive one from the same institution and batch

Repeat the measurement with an internal ruler added to every tube, so that 'there is less of it' can be told apart from 'it is somewhere else'. Run the odd patient and a normal one side by side.

The same thing, for a scientist

Every number in GSE252378 is a within-sample share because the deposit has **no spike-in**, and a share cannot see a global change - it flattens toward 1.0 when occupancy falls everywhere. That single limitation is what makes the dosage and redirection models indistinguishable in the deposited data. The experiment is therefore not a new idea but a calibration. **Design.** Two primary KMT2A-rearranged blasts processed in one batch, on one day, by one operator: one HOXA-negative (the A73051 phenotype) and one HOXA-positive, matched for institution, lineage and partner where possible. Four channels per sample - KMT2A-N1, KMT2A-N2, KMT2A-C1, KMT2A-C2 - plus IgG, plus a defined heterologous spike-in nucleosome or yeast chromatin at a fixed cell-number ratio. **Report absolute spike-in-normalised coverage, not shares.** **Readouts, in order of decisiveness.** (1) Total spike-in-normalised genome-wide N-terminal coverage, HOXA-negative versus HOXA-positive. (2) Genome-wide N-over-C scan at the authors' own thresholds, and whether the HOXA-negative sample's excess intervals form any coherent signature. (3) Fusion-junction-spanning RNA or protein quantification from the same material, so 'low on chromatin' can be separated from 'low in the cell'. (4) Matched ex vivo menin-inhibitor sensitivity.

Go deeper
What has to be true for this to be worth doing

The compute-only falsifier comes first and is free: run the genome-wide N-over-C scan on A73051's four public bigWigs and ask whether its N-excess intervals name any programme. If they do, the hypothesis is dead and no wet work is needed. If they do not, the wet experiment is the only way to distinguish low dosage from an occupancy the instrument cannot see.

Feasibility is the hard part, not the protocol. A HOXA-negative KMT2A-rearranged blast is one sample in 18 depth-usable non-AFF1 samples in the largest such deposit, so prospective collection means a multi-site banking effort. AutoCUT&RUN is already established at all four contributing institutions, which is the strongest argument for doing it as a spike-in addendum to an existing pipeline rather than as a new study.

Facts and figures
designpaired KMT2A-N and KMT2A-C AutoCUT&RUN with a defined heterologous spike-in at a fixed cell-number ratio, in one HOXA-negative and one HOXA-positive primary KMT2A-rearranged blast, same batch, same operator, same day; IgG in both; blast fraction measured in the same aliquot
readoutabsolute spike-in-normalised genome-wide N-terminal coverage; genome-wide N-over-C scan at log-fold-change 1.75 and p < 1e-5; fusion-junction-spanning transcript or protein quantification; matched ex vivo menin-inhibitor IC50
controlsIgG per sample; the eight-locus non-expressed panel (ALB, RHO, MYH7, MYOD1, NEUROD1, KRT14, GFAP, INS) as an absolute floor; a HOXA-positive blast in the same batch as the biological comparator; the spike-in itself as the between-sample ruler
cost tierhigh
timelinethe compute-only falsifier is days on public data; the wet experiment depends on prospective banking of a HOXA-negative blast, which is roughly 1 in 18 depth-usable non-AFF1 KMT2A-rearranged samples, so 12-24 months across four sites
cheaper first stepgenome-wide N-over-C scan on A73051's four public bigWigs; if its N-excess intervals name a coherent programme the hypothesis is refuted for free
statusplanned
measurements
[
  {
    "quantity": "prevalence_of_the_target_phenotype",
    "subject": "depth-usable non-AFF1 GSE252378 samples that are HOXA-negative",
    "value": 1,
    "units": "samples",
    "n": 18,
    "universe": "the 18 depth-usable non-AFF1 KMT2A-fusion samples in GSE252378",
    "null": null,
    "note": "5.6%; sets the prospective collection burden for this experiment"
  }
]
How this could bite us back
  • A spike-in-normalised negative would be strong evidence that these patients have little chromatin-bound oncoprotein, which argues against giving them a menin inhibitor - a conclusion that would be drawn from n = 1 or n = 2 unless the collection succeeds at scale.
  • Measuring occupancy to decide therapy requires a clinical-grade assay, and CUT&RUN on primary blasts is not one. A result here would create a demand the field cannot currently meet.
Still open
  1. Is a heterologous nucleosome spike-in adequate for AutoCUT&RUN on primary blasts, or does the automated protocol's bead chemistry make a cell-based spike-in the only workable ruler?
  2. Can blast fraction be measured in the same aliquot without consuming material the assay needs? If not, the spike-in normalisation inherits an unmeasured denominator.
Every source behind this entry (2)
  1. E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.dataset

    GSE252378 contains no spike-in, so every published and re-analysed quantity from it is a within-sample share that cannot detect a global change in occupancy.

    What this evidence cannot show

    The authors' own normalisation divides by total base-pair coverage of the fragment file with a pseudocount, which is also a within-library share, so this limitation is shared by their figures and by every re-analysis including this one. It is not a criticism of their work.

    Evidence that argues against this (1)
    • A spike-in is itself only as good as the cell-counting that sets its ratio, and in primary patient material with variable blast fraction the cell count is the dominant error. A spike-in would replace one uncalibrated number with a differently uncalibrated one unless blast fraction is measured in the same aliquot. Janssens DH 2024
  2. E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.dataset

    The deposit's own fusion-specificity machinery is a paired N-terminal and C-terminal comparison at log-fold-change 1.75, p < 1e-5, so the proposed scan uses the authors' published thresholds rather than new ones.

    DOI
    What this evidence cannot show

    Read from the notebook's Count_N_C function signature rather than from a methods paragraph; the exact statistical model behind the p-value was not re-derived.

    Evidence that argues against this (1)
    • Applying a published threshold to a sample with much less signal is not threshold-neutral: a fixed log-fold-change cutoff on a low-coverage library returns fewer intervals for reasons of power, so a null scan in A73051 would need a power analysis before it could be read as redirection-negative. Q148 (Oncofusion Knowledge Graph) 2026
Connected to