The MYC contradiction can be advanced but not closed: the enhancer is independently validated, the SWI/SNF route is not, and the two papers never measured the same thing

Bahr et al. 2 is not a replication of Shi et al. and should not be presented as one. It shares the element and nothing else: the perturbation is germline deletion rather than subunit knockdown, the question is Myc regulation in the normal hierarchy rather than a chromatin dependency, and the leukaemia arm is a maintenance experiment in mice. It is therefore evidence about the element, not about SWI/SNF.

What that buys is real but bounded. Before this entry, a reader of mechanism-baf-myc-contradiction-two-papers-one-model could reasonably conclude that the MYC-enhancer story in MLL-AF9 rested entirely on one paper whose central claim had failed to replicate. It does not. The element is load-bearing in MLL-AF9 leukaemia by an independent genetic test. That changes what a sceptic has to argue: not 'the MYC enhancer story is unreplicated' but 'the element matters and SWI/SNF may not be what acts there'.

And the species asymmetry is not resolved by it either, because Bahr et al.'s leukaemia genetics are also mouse. The human arm of this entire question is one signature experiment from 2015.

The NrasG12D confound. Shi et al. use RN2, an MLL-AF9;NrasG12D model. RAS signalling stabilises MYC protein, so a MYC dependency in RN2 could be partly a RAS phenotype. mechanism-baf-myc-contradiction-two-papers-one-model already lists this as an open question. This entry adds a data point that cuts both ways: Rathert et al. 5 use the same MLL-AF9;NrasG12D background and find that MYC is the node through which BET-inhibitor resistance operates, with a focal MYC enhancer recruiting WNT machinery. That is a second, independent demonstration of MYC centrality in that genotype - which strengthens the MYC claim and simultaneously strengthens the suspicion that the genotype is doing the work.

Enhancer plasticity as a reconciliation. If MYC enhancer usage re-routes readily 5, then a subunit knockdown that takes days - Cruickshank et al.'s timescale - allows compensation, whereas Shi et al.'s occupancy and looping measurements are of the unperturbed state. Under this reading both papers are correct: SWI/SNF holds the enhancer, and removing SWI/SNF slowly permits re-routing so that a steady-state MYC signature is unchanged. This is a reconciliation hypothesis, not a finding, and it predicts that an acute degron-based SMARCA4 depletion in a human KMT2A-rearranged line would show a transient MYC fall that a shRNA experiment would miss. Nobody has run it.

The contradiction is advanced and not resolved. Recorded as a disputed-adjacent state: the entry's data.status says narrowed, and the specific residual question is named. Three reasons it cannot be closed from the literature as it stands:

1. No human enhancer measurement exists. Every occupancy measurement at this element in a KMT2A-rearranged context is mouse.
2. No acute-depletion experiment exists. Every SWI/SNF perturbation in the dispute is shRNA or knockdown over days, which cannot distinguish a real independence from a compensated dependence.
3. No MYC protein measurement exists in the human lines. Both papers report transcript or signature. Under the menin-side model B in mechanism-myc-after-menin-inhibition-direct-or-differentiation-readout, transcript and protein can move independently, and the same applies here.

All three are single experiments. The coordinator's brief asked whether a third paper, a different system or a different measurement separates the two, and the answer is: a third paper strengthens the element, no system separates the papers, and the measurement that would is a BRG1 CUT&RUN at BENC in a human line that nobody has published.