Oncofusion

start / graph / ana-q130-a-replicate-pair-was-already-binned-on-disk-and-all-four-murine-vtp-contrasts-give-spearman-minus-one

Analysismoderate confidence 3 of 5: One strong study, or several smaller consistent ones.Q130-does-the-eviction-gradient-exist-in-h3k79me2

All four murine VTP-50469 H3K79me2 contrasts in GSE260742 give Spearman -1.000 against eviction depth, and all four SGC0946 contrasts are flat: the sign reproduces four times out of four, the span only to within a factor of 1.99, and the replicate pair that establishes this was already binned on disk from cycle 82

also called Q130 DEVIATION 4 · the replicate that was on disk · four of four · how well does a gradient span reproduce

A second copy of one of the experiments was already sitting on this machine from an earlier round of work. Running it gives the same direction of effect - four times out of four across all the available experiments - but the size of the effect varies by about a factor of two between copies.

The same thing, for a scientist

**Found while writing the absence entry that was going to say no replicate exists.** GSE260742 deposits 30 H3K79me2-titled samples. Two of them are a pair: GSM8123372/373/374 titled *MLLAF9 Control - DMSO / SGC0946 5uM / VTP50469 500nM H3K79me2* and GSM8123386/387/388 titled the same with **rep2** appended, with identical `cell line: MLLAF9`, `tissue: Blood`, `cell type: Myeloid` and identical exposures. **Both were downloaded and binned by Q122 in cycle 82 at the same 2 kb mm10 grid with the same `Q117Bin`, and both TSVs are still in `C:/Temp/q122`. Zero downloads** [E1]. **Every murine VTP-50469 H3K79me2 contrast that exists in this deposit** [E2]: | culture | VTP sample | Spearman vs eviction depth | span | |---|---|---:|---:| | MLLAF9 Control rep1 | GSM8123374 | **-1.000** | -0.2438 | | MLLAF9 Control rep2 | GSM8123388 | **-1.000** | -0.4847 | | NTvKO Control | GSM8123440 | **-1.000** | -0.6005 | | NTvKO PCGF1-KO | GSM8123437 | **-1.000** | -0.4341 | **And every SGC0946 contrast, the same four cultures, the same bins** [E2]: | culture | span | |---|---:| | MLLAF9 Control rep1 | +0.0031 | | MLLAF9 Control rep2 | +0.0088 | | NTvKO Control | +0.0036 | | NTvKO PCGF1-KO | +0.0109 | **The sign reproduces four times out of four. The span does not reproduce well: 0.2438 against 0.4847 between two replicates of ONE culture is a factor of 1.99.** That factor is the honest error bar on every span this beat quotes, and it is quoted in the instrument caveat of every entry [E2]. **What this arm buys and what it costs.** It buys the only replicate estimate available for H3K79me2 anywhere in this deposit. **It costs the third-culture argument**: `S_FLAG` is defined in GSM8123359, titled *MLLAF9 Control - DMSO FLAG* - **the same culture as this replicate pair** - so for these two contrasts the stratifier and the measurement are not library-independent. The NTvKO contrasts have the independence and no replicate; these have the replicate and no independence. **Neither alone is sufficient and that is why all four are reported** [E3].

Go deeper
What a 1.99-fold span reproducibility does to the rest of the beat

It does not touch any sign claim. Four of four at Spearman -1.000, and eight of eight if the four flat SGC contrasts are counted as the negative.

It bounds every magnitude claim to about a factor of two. The headline's 0.6005 is the largest of the four spans and the median is about -0.459, so a reader should hold 'about a half' rather than '0.6005'. The entries quote the NTvKO number because it is the library-independent one AND the one Q124 scored, but the objection is recorded in the headline's own counter-evidence.

And it puts a floor under the drug-discrimination claim rather than a ceiling. The smallest VTP span in the deposit is 0.2438 and the largest SGC span is 0.0109 - a 22-fold separation with no overlap across eight contrasts - so a 2-fold reproducibility in the VTP spans cannot close a 22-fold gap.

Facts and figures
scriptanalyses/q130-does-the-eviction-gradient-exist-in-h3k79me2/deviation4_the_replicate_that_was_on_disk.py
statusrun
result summaryFour of four murine VTP contrasts at Spearman -1.000, spans -0.2438 / -0.4847 / -0.6005 / -0.4341; four of four SGC contrasts flat at +0.0031 to +0.0109; span reproducibility 1.99-fold between two replicates.
downloads required0
absence classmeasured-not-analysed
absence class noteThe replicate pair was binned in cycle 82 and has been on disk for eleven cycles. Q122 scored it as aggregates. Nobody stratified it, and this beat's own pre-registration stated - wrongly - that no replicate exists and that library-to-library variation was unmeasurable. The correction is recorded in the instrument caveat of every entry in the beat rather than only here.
absence class assigned2026-09-23 (Q130)
epigenetic readoutH3K79me2 eviction-depth span, reproducibility across four murine cultures
per contrast
REP1_VTP-1.000REP1_VTP: -1REP2_VTP-1.000REP2_VTP: -1REP1_SGC0.200REP1_SGC: 0.2REP2_SGC0.800REP2_SGC: 0.8
inputs
[
  "GSM8123372/373/374 MLLAF9 Control (binned by Q122, cycle 82)",
  "GSM8123386/387/388 MLLAF9 Control rep2 (binned by Q122, cycle 82)",
  "GSM8123435-440 NTvKO (binned by Q124, cycle 83)"
]
reproducibility
{
  "rep1_span": -0.2438,
  "rep2_span": -0.4847,
  "absolute_difference": 0.2409,
  "ratio_larger_over_smaller": 1.988,
  "all_four_VTP_spans_in_this_deposit": {
    "MLLAF9 Control rep1 (GSM8123374)": -0.2438,
    "MLLAF9 Control rep2 (GSM8123388)": -0.4847,
    "NTvKO Control (GSM8123440)": -0.6005,
    "NTvKO PCGF1-KO (GSM8123437)": -0.4341
  },
  "all_four_VTP_spearmans": {
    "MLLAF9 Control rep1": -1,
    "MLLAF9 Control rep2": -1,
    "NTvKO Control": -1,
    "NTvKO PCGF1-KO": -1
  },
  "all_four_SGC_spans": {
    "MLLAF9 Control rep1": 0.0031,
    "MLLAF9 Control rep2": 0.0088,
    "NTvKO Control": 0.0036,
    "NTvKO PCGF1-KO": 0.0109
  }
}
per contrast — the numbers behind the chart
{
  "REP1_VTP": {
    "R_by_quartile": [
      0.6405,
      0.7784,
      0.8146,
      0.8843
    ],
    "spearman_vs_depth": -1,
    "span_Q1_minus_Q4": -0.2438,
    "R_aggregate_S_FLAG": 0.7736,
    "median_per_bin_log2_by_quartile": [
      -0.5749,
      -0.3559,
      -0.3037,
      -0.1634
    ],
    "bootstrap_span_ci95": [
      -0.2877,
      -0.2001
    ]
  },
  "REP1_SGC": {
    "R_by_quartile": [
      0.0926,
      0.096,
      0.0976,
      0.0895
    ],
    "spearman_vs_depth": 0.2,
    "span_Q1_minus_Q4": 0.0031,
    "R_aggregate_S_FLAG": 0.0939,
    "median_per_bin_log2_by_quartile": [
      -3.4348,
      -3.2734,
      -3.3005,
      -3.3397
    ],
    "bootstrap_span_ci95": [
      -0.0059,
      0.0116
    ]
  },
  "REP2_VTP": {
    "R_by_quartile": [
      0.6383,
      0.8401,
      0.9142,
      1.123
    ],
    "spearman_vs_depth": -1,
    "span_Q1_minus_Q4": -0.4847,
    "R_aggregate_S_FLAG": 0.8324,
    "median_per_bin_log2_by_quartile": [
      -0.5671,
      -0.1653,
      -0.0623,
      0.1637
    ],
    "bootstrap_span_ci95": [
      -0.5351,
      -0.435
    ]
  },
  "REP2_SGC": {
    "R_by_quartile": [
      0.1531,
      0.1533,
      0.1483,
      0.1443
    ],
    "spearman_vs_depth": 0.8,
    "span_Q1_minus_Q4": 0.0088,
    "R_aggregate_S_FLAG": 0.1506,
    "median_per_bin_log2_by_quartile": [
      -2.7067,
      -2.6702,
      -2.6882,
      -2.6923
    ],
    "bootstrap_span_ci95": [
      -0.0033,
      0.0204
    ]
  }
}
How this could bite us back
  • A pharmacodynamic statistic whose magnitude reproduces only to within a factor of two cannot be used to grade depth of response between patients. It can distinguish drug from no drug and menin inhibition from DOT1L inhibition, and that is all this reproducibility supports.
Still open
  1. Does the H2AK119ub / H3K27me3 / H3K4me3 gradient reproduce as well? Q125's marks have no replicate in the NTvKO arm either, and the MLLAF9 Control replicate pair carries H2AK119ub too (Q125 binned both conventions) - so the same four-of-four test is available for one more mark at zero cost.
  2. Is the 1.99-fold span difference technical or between-culture? It depends on resolving whether GSM8123386-388 is rep2 or a separate F3_NT culture, which the companion dispute entry reports and does not settle.
Every source behind this entry (3)
  1. E1strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.dataset

    GSE260742 deposits a replicate pair for the MLLAF9 Control culture's H3K79me2 arm - GSM8123372/373/374 and GSM8123386/387/388, the latter titled 'rep2' with identical cell line, tissue, cell type and exposures - and both were binned by Q122 in cycle 82 at the same 2 kb mm10 grid, so scoring them cost zero downloads.

    What this evidence cannot show

    THE PAIR'S STATUS IS DISPUTED AND THE DISPUTE IS REPORTED IN A COMPANION ENTRY: the rep2 samples' GEO titles say 'rep2' while their supplementary filenames say 'ChIP_K79_F3_NT_*'. Q122 named them F3NT and Q124 treats them as a separate culture. Either reading supports the measurement here - a replicate or a second culture - which is why it is reported before the dispute is settled. The rep1 files are .filt.bam.tdf and the rep2 files are bare .tdf, so the pair is not a pure technical replicate, although the rep2 TDF's own internal track name reads 'ChIP_K79_F3_NT_VTP.filt.bam', i.e. the same filtered BAM lineage.

    Evidence that argues against this (1)
    • If the pair is two cultures rather than two replicates, the 1.99-fold span difference is biological variation between cultures and is not a technical error bar at all. Iqbal Lab oncofusion knowledge graph 2026
  2. E2moderate confidence 3 of 5: One strong study, or several smaller consistent ones.dataset

    All four murine VTP-50469 H3K79me2 contrasts in GSE260742 give Spearman -1.000 against eviction depth with spans -0.2438, -0.4847, -0.6005 and -0.4341, and all four matching SGC0946 contrasts are flat with spans +0.0031, +0.0088, +0.0036 and +0.0109.

    What this evidence cannot show

    NO SPIKE-IN NORMALISATION IS POSSIBLE. Q117 verified zero hits for the entire spike-in vocabulary across all 419,824 characters of GSE260742's 158 sample records and no scale factor is deposited, so every cross-condition number here is a WITHIN-SAMPLE SHARE. It cancels the unknown per-sample scale factor exactly and IS BLIND TO A UNIFORM CHANGE BY CONSTRUCTION: nothing here measures H3K79me2's abundance, everything measures its distribution. EXPOSURE ASYMMETRY: DMSO 72 h and SGC0946 5 uM 72 h are matched, VTP-50469 500 nM is 48 h, so every VTP ratio is a LOWER BOUND on the 72-hour effect - identical in both genotypes, so the genotype DIFFERENCE is unaffected. REPLICATION, CORRECTED MID-BEAT: the NTvKO arm deposits ONE H3K79me2 ChIP per genotype per drug, so ITS spans carry no technical error bar and its bin-level bootstrap intervals measure sampling within one library, NOT library-to-library variation. A REPLICATE PAIR DOES EXIST for a different culture - GSM8123372/374 and GSM8123386/388, both titled MLLAF9 Control, the second carrying `rep2` - already binned by Q122 and scored in Q130 DEVIATION 4: the SIGN and the monotonicity reproduce (Spearman -1.000 in all four murine VTP contrasts in this deposit) while the SPAN reproduces only to within a factor of 1.99 (-0.2438 against -0.4847). EVERY SPAN IN THIS BEAT SHOULD BE READ WITH THAT 2-FOLD REPRODUCIBILITY. THE KNOCKOUT is one CRISPR guide, one clone C1, no second clone, no rescue, and NO deposited PCGF1 protein or transcript measurement. S_FLAG is defined by a DYKDDDDK-tag ChIP on a RETROVIRALLY OVER-EXPRESSED tagged MLLAF9 construct, so it is the over-expressed molecule's sites and not an endogenous fusion's. THE H3K79me2 ANTIBODY IS NOT RESOLVED - `antibod` occurs zero times in the 158 GEO records - and is recorded unresolved rather than guessed. SYSTEM: murine MLL-AF9 (KMT2A::MLLT3) myeloid blood cells, NOT the infant B-lineage KMT2A::AFF1 disease this graph is about.

    Evidence that argues against this (2)
    • A published methods study establishes that STANDARD ChIP-seq normalisation FAILS TO DETECT a genome-wide decrease in a histone mark under an inhibitor of its writer: standard methods see no H3K27me3 reduction under an EZH2 inhibitor, and a substantial reduction appears only with a Drosophila chromatin spike-in. Every number in this entry is exactly the kind of normalisation that paper shows to be blind. Egan B et al. 2016
    • The system is murine MLL-AF9 (KMT2A::MLLT3) myeloid cells carrying a retrovirally over-expressed FLAG-tagged construct. This graph exists for human infant B-lineage KMT2A::AFF1, and this corpus's commonest documented synthesis failure is generalising a murine MLL-AF9 myeloid result to that disease. Neville D et al. 2026
  3. E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.dataset

    The third-culture independence argument does NOT apply to the MLLAF9 Control replicate pair, because S_FLAG is defined in GSM8123359 'MLLAF9 Control - DMSO FLAG', the same culture; the NTvKO contrasts have the independence and no replicate, and the MLLAF9 Control contrasts have the replicate and no independence.

    What this evidence cannot show

    This is a structural statement about which libraries define and which measure, read from the samples' own titles. It is not a measurement and cannot be quantified; its value is in preventing the four contrasts from being quoted as four equivalent confirmations.

    Evidence that argues against this (1)
Connected to