Measured in six ENCODE cell types against every conservative-IDR CTCF anchor of the same cell: the KMT2A exon-11 anchor sits at the 70.6th percentile of CTCF occupancy (range 28.2-81.8, fold over the anchor median 0.69-2.37) - a called peak in 6 of 6 cells and an ORDINARY one, which fails the instrument control the corpus pre-registration wrote for this very experiment

**What was measured and why.** The cycle-64 experiment `exp-idr-free-signal-level-reread-of-the-kmt2a-anchor-across-every-archived-cbx3-and-brca1-track` pre-registered an instrument control, **C4**: *"a CTCF track of the same cell type must place the exon-11 anchor above its own 90th percentile, or the pipeline cannot see this element in that cell."* C4 had never been run. It was run here first, before any CBX3 or BRCA1 target track was read, exactly so that a blind cell type would be known before its targets were scored [E1]. **C4 FAILS IN 6 OF 6 CELL TYPES, and the reason is not blindness.** The anchor is a *called* conservative-IDR CTCF peak in every one of the six ENCODE cell types with a CTCF peak set, one peak per cell, md5-verified against the portal record. Its fold-change signal over a 147 bp window centred on its own narrowPeak point-source summit, ranked against every other conservative-IDR CTCF anchor of the same cell (42,947-54,654 anchors, a census and not a sample), is: | cell | CTCF experiment | signal file | n anchors | fold over anchor median | percentile | C4 (>90th) | |---|---|---|---:|---:|---:|---| | HCT116 | ENCSR000DTO | ENCFF388PVO | 49,555 | 2.372 | **81.76** | FAIL | | HepG2 | ENCSR000DLS | ENCFF579BUC | 42,947 | 1.910 | **79.98** | FAIL | | HeLa-S3 | ENCSR000AOA | ENCFF179RSE | 45,978 | 2.022 | **78.67** | FAIL | | K562 | ENCSR000EGM | ENCFF682MFJ | 54,654 | 1.326 | **62.43** | FAIL | | H1 | ENCSR000BNH | ENCFF038RVZ | 45,570 | 0.978 | **48.44** | FAIL | | GM12878 | ENCSR000DZN | ENCFF485CGE | 44,217 | 0.691 | **28.15** | FAIL | Median **70.61**, range 28.15-81.76, **0 of 6 above the 90th percentile** [E1]. **So the pre-registration's premise was wrong, not its arithmetic.** C4 was written on the assumption that this element is a top-decile CTCF anchor; it is not. The element is detected everywhere (a called conservative-IDR peak in 6/6) and is ordinary everywhere. Reading C4's failure as "the pipeline is blind" would be the error; the honest reading is that **the pre-registered control measured the anchor's rank and found it middling**, which is a result about the anchor rather than about the instrument, and which this entry records rather than letting the threshold be quietly moved. **This is a different answer from the one this graph already holds, in a different cell type.** `genome-kmt2a-exon-11-ctcf-anchor-ranked-among-matched-rpe-anchors` places the same element at the **91.7th percentile** of 12,443 de-novo-called RPE anchors from GSE210480. That set is a quarter the size of ENCODE's and is truncated from below at 2.51x the track mean, so its median anchor is *stronger*, which should push a fixed element's percentile **down**, not up. The two readouts therefore disagree in the direction that cannot be explained by anchor-set size, and the residual candidates are cell type (RPE vs six cancer/stem lines - and this run's own 28-82 spread shows how cell-type-variable the rank is) and track normalisation (HOMER-derived ratio track vs ENCODE fold-change-over-control). **Neither number is withdrawn; the disagreement is the finding, and it means 'how strong an anchor is this' has no single answer.** [E1][E2] **A second, cleaner discordance inside the same experiment.** The narrowPeak `signalValue` of the KMT2A anchor ranks at the **94.7th** percentile of GM12878's own CTCF peaks and at the **91.0th** in K562 - while the fold-change bigWig from the *same ENCODE analysis* (verified: ENCFF485CGE and ENCFF796WRU are both files of `/analyses/ENCAN431GAG/`, ENCODE4 v1.5.1 GRCh38) puts it at the **28.2nd** and **62.4th**. A 66.5-point gap between two readouts of one experiment. Per-cell signalValue percentiles: GM12878 94.7, K562 91.0, HepG2 88.1, HCT116 78.4, HeLa-S3 75.1, H1 70.2 [E1].