The menin interface is identical in all 27 of these patients and the amount of chromatin their oncoproteins occupy differs by up to two orders of magnitude - so the interface can be a permission for binding but cannot be what sets how much binding happens

**The constraint, stated as a bound.** The menin-binding motifs MBM1 and MBM2 lie in the retained KMT2A N-terminus and are present, intact and identical in **every** KMT2A fusion in this cohort - AFF1, ENL, AFDN, AF10, AF9 and EPS15 alike, since all six break downstream of the menin-binding region. Across those same 27 samples the number of promoters passing the published oncoprotein criterion runs **%d to %d**, with a %sx spread between the 10th and 90th percentiles and a %sx ratio between the largest and smallest non-zero count, and the spread **widens** after depth matching [E1]. **Therefore an invariant interface is compatible with a hundred-fold range in occupancy.** The interface cannot be the quantitative determinant. This is the amount-side twin of the corpus's existing address-side constraint, which was derived from one patient's two disease states: an unchanged interface is compatible with an almost completely changed target set [E2]. **Together the two say the interface sets neither WHERE the fusion goes nor HOW MUCH of it goes there.** **What the interface can still be, and this is not a weakening of the drug rationale.** A necessary permission. Menin inhibition removes a required contact and the fusion leaves chromatin - which this beat's own positive control shows at the extreme, with the post-menin-inhibitor aliquot the only one of 27 calling zero promoters [E3]. A gate that must be open for anything to happen is fully compatible with something else deciding how much happens once it is open. **What would refute this constraint.** A demonstration that the between-patient occupancy range is driven by *menin abundance* rather than by the fusion. Menin concentration varies between cells, and a fusion whose recruitment is menin-limited would show exactly this spread with an invariant interface - the interface would then be the mechanism and menin the dial. **This deposit cannot test that**, because it contains menin CUT&Tag in only one patient's two aliquots [E1]. A menin channel in the other 26 would settle it, and it is the single most valuable missing measurement identified by this beat. **And the alternative that must be excluded first.** Tumour purity. A marrow with few blasts dilutes the fusion signal with normal cells whose N/C ratio is near one, which lowers the call count with no change in the leukaemic cells. **The deposit records no blast percentage**, so the ranking of patients by occupancy is not yet separable from a ranking by purity, and this mechanism entry inherits that uncertainty.