Scored in the published criterion's own currency, the number of oncoprotein-target promoters runs 0 to 2282 across 27 KMT2A-rearranged patient samples - 20.01x between the 10th and 90th percentiles - and the spread widens to 61.10x after depth matching

**The statistic.** Per sample, the number of TSS +/- 1 kb promoters passing the paper's own criterion - **N/C ratio > 1.75 with binomial upper-tail p < 1e-5** against the within-sample null share TN/(TN+TC). It runs from **0 to 4646** by construction, so it is unbounded in the sense this beat was dispatched to supply [E1]. **Measured across 27 samples: 0 to 2282 called promoters, median 947.00. One sample calls ZERO. The robust spread (p90/p10) is 20.01x and the largest over the smallest NON-ZERO count is 103.73x** [E1]. **It is not a depth artefact, and that was checked rather than assumed.** A binomial threshold has more power in a deeper library, so this was the first thing to break. Spearman of the call count against log10 total fragments is **-0.165 (p = 0.411)**. Thinning every library by seeded binomial down-sampling to a common **6,811,300** fragments per arm leaves the spread at **0 to 1802, p90/p10 = 61.10x** - wider, not narrower [E2]. **This is the magnitude answer the corpus asked for, and its sign is unambiguous.** Whatever else is true, the amount of chromatin a KMT2A oncoprotein occupies, measured the way the people who built the assay measure it, is not a constant of the fusion. It is a property of the patient, and it varies by two orders of magnitude. **What it does NOT say.** There is **no spike-in in this deposit**. Every input is a within-sample share, so this is a statement about how much of a library's signal is N-biased, not about how many oncoprotein molecules are on chromatin per cell. A patient with fewer called promoters may have less oncoprotein, or the same oncoprotein concentrated on fewer sites, and nothing here separates those.