Oncofusion

start / graph / ana-q118-instrument-check-the-top-rs4-11-mll1-n-bins-are-jmjd1c-cdk6-myb-runx2-and-igf2bp2

Analysisstrong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.Q118-menin-at-the-human-index-fusion

Instrument check: the top thirty RS4;11 MLL1-N 2 kb bins recovered from a deposited .tdf are JMJD1C (259x the genome median), ZMYND11, BAHCC1, CDK6, MYB, RUNX2, IGF2BP2, FOXP1, SATB1 and ARID1B - canonical KMT2A::AFF1 targets, annotated only against the series' own RefGene windows

also called Q118 instrument check · the top bins are the right genes

Before trusting any measurement, the strongest signals in the re-processed data were listed and checked against what is already known to be bound in this leukaemia. They are the expected genes, which means the pipeline that read the deposited files is working.

The same thing, for a scientist

**Before any verdict, the instrument was checked the way Q117 checked its own: list the top bins and read them.** The top thirty RS4;11 MLL1-N DMSO 2 kb bins (universe 1,389,514 bins with input > 0 and not blacklisted), with the enrichment over the track's own genome median, and the within-sample share ratio under VTP-50469 for both antibodies at each bin [E1]: | locus (hg19) | enrich | MLL1-N ratio | MENIN ratio | ROSE window | |---|---:|---:|---:|---| | chr10:65,224,000 | **259.0x** | 0.138 | 0.031 | JMJD1C-AS1 | | chr4:15,004,000 | 199.5x | 0.385 | 0.124 | LOC441009 | | chr10:180,000 | 183.0x | 0.133 | 0.052 | ZMYND11 | | chr15:28,982,000 | 164.1x | 0.988 | 0.105 | WHAMMP2 | | chr10:65,280,000 | 150.1x | 0.091 | 0.037 | REEP3 | | chr17:79,370,000 | 128.3x | **4.036** | 0.056 | BAHCC1 | | chr9:80,646,000 | 125.2x | 0.309 | 0.071 | GNAQ | | chr3:71,630,000 | 105.2x | 0.509 | 0.052 | FOXP1 | | chr3:151,986,000 | 104.9x | 0.640 | 0.081 | MBNL1 | | chr7:92,462,000 | 99.5x | 0.222 | 0.045 | **CDK6** | | chr3:185,542,000 | 99.1x | 0.177 | 0.070 | IGF2BP2 | | chr6:135,502,000 | 94.8x | **1.254** | 0.076 | **MYB** | | chr6:45,390,000 | 76.0x | 0.286 | 0.030 | **RUNX2** | | chr6:157,100,000 | 75.8x | 0.236 | 0.072 | ARID1B | | chr3:18,466,000 | 73.4x | 0.618 | 0.150 | SATB1 | **JMJD1C, CDK6, MYB, RUNX2, IGF2BP2, FOXP1, ARID1B, SATB1, ZMYND11, BAHCC1** - these are the KMT2A::AFF1 target set, and the pipeline found them without being told to look. **76.1% of the 1,390 held bins fall inside one of the deposited series' own 21,461 RefGene TSS windows, against 6.4% of the universe.** **And the per-bin MENIN column is the whole beat in one image.** At every one of the top thirty bins the MENIN share ratio is between **0.030 and 0.296** - menin is gone everywhere - while the MLL1-N ratio ranges from **0.091 to 4.036**. Two adjacent columns, same bins, same cells. **Two bins deserve naming.** chr17:79,370,000 and chr17:79,314,000 at **BAHCC1** carry MLL1-N ratios of **4.036** and **3.407** - the largest per-bin rises in the top thirty - with menin at 0.056 and 0.078. **BAHCC1 is an H3K27me3 reader implicated in leukaemia and it is not a locus this graph has an entry for.** It is the single most interesting unexplained observation in this beat. **Annotation provenance:** the gene names above come **only** from overlap with the series' own 6 kb ROSE RefGene TSS windows. **No external gene-coordinate resolver was used anywhere in this beat**, deliberately, because Q117 lost time to the UCSC `search` API placing Hoxa9 on chr1 under a longest-match rule. A bin in no ROSE window is reported as `-`, not guessed.

Facts and figures
scriptanalyses/q118-menin-at-the-human-index-fusion/binsD.py (D8 block)
statusrun
result summaryTop thirty RS4;11 MLL1-N bins recover JMJD1C, ZMYND11, BAHCC1, CDK6, MYB, RUNX2, IGF2BP2, FOXP1, SATB1, ARID1B and CDKN1B; 76.1% of held bins are at a RefGene TSS against 6.4% of the universe; MENIN is 0.030-0.296 at all thirty while MLL1-N is 0.091-4.036, peaking at BAHCC1.
cell lineRS4;11
fusionKMT2A::AFF1
genome buildhg19
bin bp2000
annotation provenanceoverlap with the deposited series' own 21,461 ROSE RefGene 6 kb TSS windows; NO external gene-coordinate resolver was used
shown for fusionKMT2A::AFF1 in human RS4;11 via an MLL1 N-terminus antibody that also sees wild-type MLL1
inputs
[
  "GSM3635695/96 RS4;11 MLL1n",
  "GSM3635693/94 RS4;11 MENIN",
  "GSM3635699 RS4;11 input",
  "GSE127507 ROSE TSS windows as the only annotation"
]
CORRECTED CYCLE 79 one of the two bahcc1 bins is not at bahcc1
{
  "what_this_entry_said": "that chr17:79,370,000 and chr17:79,314,000 are both 'at BAHCC1'",
  "verified_live_by_king": "Ensembl GRCh37 overlap API, region 17:79300000-79450000, feature=gene, fetched and read 2026-09-23. BAHCC1's GRCh37 model is ENSG00000171282 at 79,373,540-79,433,357.",
  "chr17_79_370_000": "3,540 bp UPSTREAM of BAHCC1's start and inside the overlapping protein_coding model ENSG00000263053 (79,369,275-79,375,658). Promoter-proximal, so 'at BAHCC1' is defensible for this one.",
  "chr17_79_314_000": "OVERLAPS ZERO GENE MODELS OF ANY BIOTYPE. 59,540 bp upstream of BAHCC1's start; nearest features are TMEM105 ending 79,304,474 (9.5 kb away) and the lincRNA RP11-1055B8.2 starting 79,336,072 (22 kb away). It is in a gene desert and IS NOT AT BAHCC1.",
  "the_corrected_wording": "an unannotated candidate distal element ~59.5 kb upstream of BAHCC1",
  "this_entry_ALREADY_CONTAINED_ITS_OWN_REFUTATION": "its own open_questions field says 'four of the top thirty bins are in no RefGene TSS window (... chr17:79,314,000)'. The correct observation was four fields away from the wrong one, in the same file, and the wrong one is what a reader would quote.",
  "why_the_fix_makes_it_more_interesting": "a reproducible high-amplitude bin in a gene desert is a candidate enhancer, which is a better finding than a second promoter-proximal bin at a gene that already had one.",
  "found_by": "q119, cycle 79; coordinates re-fetched from Ensembl GRCh37 by the king"
}
Still open
  1. BAHCC1 carries the two largest per-bin MLL1-N rises in the top thirty (4.036 and 3.407) with menin falling to 0.056 and 0.078. BAHCC1 is an H3K27me3 reader with a published role in leukaemia and it has NO entry in this graph. Is the rise real, and is BAHCC1 an unwritten KMT2A::AFF1 target?
  2. The top bin at 259x is annotated JMJD1C-AS1 and the fifth at 150x is REEP3, both in the JMJD1C neighbourhood on chr10. What is the actual structure of MLL1-N occupancy across the JMJD1C locus at 2 kb resolution? The binned data on disk answers it in minutes and it was not done.
  3. Four of the top thirty bins are in no RefGene TSS window (chr2:66,672,000, chr4:180,980,000 and 180,978,000, chr17:79,314,000). High-amplitude non-promoter MLL1-N sites in the index fusion are exactly the enhancer compartment this literature discusses and they are unannotated here.
Every source behind this entry (1)
  1. E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.dataset

    The top thirty RS4;11 MLL1-N DMSO 2 kb bins reach 68.9-259.0 fold over the track's own genome median and fall in the ROSE TSS windows of JMJD1C-AS1, ZMYND11, REEP3, BAHCC1, GNAQ, FOXP1, MBNL1, CDK6, IGF2BP2, MYB, APOLD1, FUT4, RUNX2, ARID1B, EMB, CDK13, SATB1, H2AFY, KCNN2, PHF10 and CDKN1B; 76.1% of the 1,390 held bins are in a ROSE TSS window against 6.4% of the universe; and at every one of the top thirty the MENIN share ratio is 0.030-0.296 while the MLL1-N ratio is 0.091-4.036, the two largest being 4.036 and 3.407 at BAHCC1.

    What this evidence cannot show

    NEITHER ANTIBODY IN THIS SERIES IS FUSION-SPECIFIC AND THE ANTIBODY NAME SAYS SO. MLL1n is anti-MLL1 N-terminus (Bethyl A300-086A, lots 5 or 6); the N-terminus is retained in KMT2A::AFF1, so that track is the sum of the fusion and wild-type MLL1 in unknown proportion. MENIN is Bethyl A300-105A Lot 8; menin binds the fusion, wild-type MLL1 and non-KMT2A partners. Q117 could attribute its numbers because its fusion was FLAG-tagged; nothing here can. Every occupancy statement in this beat is about MLL1-N-reactive material or menin-occupied sites, never about the fusion. No scale factor exists for any sample. The depositors state that a Drosophila spike-in was run on one batch, that fly/human ratios were 'found to be small', and that NO spike-in normalisation was applied; no number for that measurement is reported and none is deposited. Every cross-condition figure here is therefore a WITHIN-SAMPLE share or a within-sample rank, which cancel an unknown per-sample scale factor exactly, and never a fold change. n = 1 per antibody per condition per line - there is no replicate anywhere in GSE127507 - so no number here carries a spread and none is precise. Gene names are by overlap with the deposited series' own 6 kb RefGene TSS windows and nothing else, so a bin can be assigned to an antisense or neighbouring gene (JMJD1C-AS1 rather than JMJD1C, REEP3 rather than JMJD1C's body) and four of the top thirty are in no window at all. The universe figure of 6.4% is from a 1-in-37 systematic sample, not a full pass. Per-bin share ratios from an n=1 comparison at single 2 kb bins are the noisiest quantity in this beat.

    Evidence that argues against this (2)
    • An instrument check that recovers the expected genes confirms the pipeline and NOT the verdict. Q117 made the same point: its instrument check passed and three of its pre-registered priors were still wrong. Recovering JMJD1C and CDK6 says the TDF reader and the binning work; it says nothing about whether the share statistic answers the question. Iqbal Lab oncofusion knowledge graph 2026
    • chr10:65,224,000 is annotated JMJD1C-AS1 and chr10:65,280,000 is annotated REEP3, and both are within the JMJD1C locus neighbourhood. Reporting the top bin as 'JMJD1C' would be a small act of interpretation the annotation does not license, and the entry states the window name instead. Iqbal Lab oncofusion knowledge graph 2026
Connected to