EHMT2 / G9a - the H3K9me1/2 writer that sets whether the KMT2A LOCUS ITSELF rearranges, and whose loss delays MLL-AF9 AML in a supplementary figure nobody can find

This is a prediction, not a measurement. No one has seen this protein in this shape. The colours say how much the software trusts itself, and that is worth reading before anything else — the orange stretches are places the model is effectively saying this part has no fixed shape.

• The direction is contested inside this graph: `mark-h3k9me3` argues that H3K9 methylation RESTRAINS the MLL-fusion programme, so an H3K9 methyltransferase inhibitor is a candidate for making a KMT2A-rearranged leukaemia worse. This entry's downstream evidence points the other way. Anyone proposing a G9a inhibitor here must first establish which mark their agent lowers, because UNC0638 lowers both.
• Tool-compound ambiguity is the practical form of that risk: UNC0638 and UNC0642 inhibit G9a and GLP together, and UNC0638 reduces H3K9me2 AND H3K9me3, so no published experiment with these compounds separates the two hypotheses.
• A catalytic inhibitor is not protein loss: the ANK repeats bind H3K9me1/2 independently of the SET domain, so G9a can continue to act as a reader and scaffold under full catalytic inhibition. Every G9a inhibitor result is therefore a statement about methyltransferase activity only, and the genetic deletions this entry rests on removed both functions.
• The upstream and downstream rationales are at different times in the disease and must not be combined: G9a inhibition suppressing KMT2A rearrangement is a PREVENTION argument in a therapy-exposed cell, and G9a loss delaying AML is a TREATMENT argument in an established leukaemia. One agent cannot inherit both without an experiment in each setting.
• Essential-function toxicity is low but the therapeutic effect is modest: G9a is not a DepMap dependency in any arm (mean -0.2260), the survival gain in the best genetic experiment was 44 days against 23.5 days in mice that still died, and the normal-tissue argument rests on mice with no overt haematological abnormality rather than on any human safety data.
• Fetal-liver-high expression cuts both ways: it makes G9a attractive as a lever in the cell of origin, and it makes an infant the worst population in which to inhibit it, because the tissue where the enzyme is most expressed is the one an infant has most recently been using.
• Compensation: EHMT1/GLP is an obligate heterodimer partner with its own catalytic SET domain and is itself a non-dependency, which is the classic signature of a redundant pair - and SETDB1 and SUV39H1/2 write the neighbouring H3K9me3, so the residue can stay methylated after G9a is removed.

1. Does the KDM3B/G9a H3K9me1/2 balance at the KMT2A locus operate in FETAL haematopoietic cells, where this disease's break actually happens? The published model is doxorubicin-driven therapy-induced rearrangement in cultured cells and mice; the index disease is de novo infant t(4;11) with an in-utero origin and no anthracycline exposure.
2. Is CTCF occupancy at the KMT2A breakpoint cluster region measured anywhere in this graph? The locus paper makes CTCF loss sufficient to produce focal rearrangement, and this project's break model is built on topoisomerase II, replication stress and nuclease cleavage with no CTCF layer at all.
3. Would G9a inhibition during anthracycline exposure reduce therapy-induced KMT2A rearrangement in a model of secondary leukaemia? That is a prophylactic hypothesis, which this graph has almost none of, and it has a named drug pair (doxorubicin plus a G9a inhibitor).
4. Does G9a support HOXA9-dependent transcription in a KMT2A::AFF1 B-lineage cell as it does in murine myeloid Hoxa9/Meis1 AML? Everything downstream here is myeloid and murine, and the one MLL-AF9 arm is a supplementary survival curve.
5. Can the G9a and SETDB1 directions be separated experimentally? Both are H3K9 methyltransferases with opposite predicted effects in this disease and the available tool compound hits both marks; a selective-degron or catalytic-dead comparison, mark-resolved by mass spectrometry rather than antibody, is the experiment.
6. Does the ANK reader module matter? A SET-domain inhibitor leaves it intact, so a degrader and an inhibitor should give different answers - and if they do, every published G9a inhibitor result in leukaemia is a statement about catalysis only.

1. E1established confidence 5 of 5: Multiple independent lines of evidence; treated as settled in the field.database

   EHMT2/G9a (Q96KQ7, 1,210 aa) mono- and dimethylates histone H3 Lys-9 in EUCHROMATIN, also monomethylates H3K56 in G1 phase and mono/dimethylates histone H1.4 Lys-26, weakly methylates H3K27, and methylates non-histone substrates including p53/TP53 at Lys-373; it heterodimerises with EHMT1/GLP, its ANK repeats BIND H3K9me1 and H3K9me2, it interacts with REST only in the presence of CDYL within a complex containing CDYL, REST, WIZ, SETDB1, EHMT1 and EHMT2, it is located in Nucleus and on Chromosome, and it is expressed in all tissues examined with HIGH levels in fetal liver, thymus, lymph node, spleen and peripheral blood leukocytes and LOWER level in bone marrow.

   UniProtKB Q96KQ7 (EHMT2_HUMAN), Histone-lysine N-methyltransferase EHMT2 - FUNCTION, CATALYTIC ACTIVITY, SUBUNIT, SUBCELLULAR LOCATION, TISSUE SPECIFICITY, DOMAIN and feature table, fetched 2026-09-22 · UniProt Consortium 2026
   ▸What this evidence cannot show

   A database record, not a measurement. The localisation half is thin in the same way PRMT6's is: both location values, Nucleus and Chromosome, rest on ONE 2001 citation (PubMed:11316813), which was not read for this entry - rule 91 applies. The tissue-specificity statement is an mRNA/protein expression survey with no method or cohort stated in the record, so 'high in fetal liver' is a database assertion and not a measurement this entry can characterise. Note that G9a is NOT annotated as an H3K9me3 writer anywhere in this record, which matters for reading it against mark-h3k9me3.

   ▸Evidence that argues against this (1)

   • The record's own architecture undercuts a simple writer reading, in a way that complicates any inhibitor argument: the ANK repeats at 649-879 BIND H3K9me1 and H3K9me2, so G9a reads the mark it writes and the protein is a self-reinforcing module rather than a one-way enzyme. A SET-domain catalytic inhibitor such as UNC0638 leaves the ANK reader intact, so it does not phenocopy protein loss - and in this graph that distinction has already mattered once, where a catalytic-dead versus degrader test was written as a dedicated experiment for a different chromatin enzyme. UniProt Consortium 2026
2. E2strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

   Histone H3 Lys-9 mono- and dimethylation balance at the MLL/KMT2A locus regulates amplification and rearrangement of that locus, and is controlled by crosstalk between the demethylase KDM3B and the methyltransferase G9a/EHMT2: KDM3B depletion increases H3K9me1/2 and reduces CTCF occupancy at the locus and promotes amplification and rearrangement, depleting CTCF alone is sufficient to generate the same focal alterations, doxorubicin suppresses KDM3B and CTCF protein levels while KDM3B or CTCF overexpression rescues doxorubicin-induced MLL/KMT2A alterations, and G9a inhibition in human cells or in mice suppresses the MLL/KMT2A events accompanying doxorubicin treatment.

   Epigenetic balance ensures mechanistic control of MLL amplification and rearrangement. · Gray ZH, Chakraborty D, Duttweiler RR, Alekbaeva GD, Murphy SE, Chetal K, Ji F, Ferman BI, Honer MA, Wang Z, Myers C, Sun R, Kaniskan HU, Toma MM, Bondarenko EA, Santoro JN, Miranda C, Dillingham ME, Tang R, Gozani O, Jin J, Skorski T, Duy C, Lee H, Sadreyev RI, Whetstine JR · Cell 2023
   PMID 37788669DOI
   ▸What this evidence cannot show

   Abstract read in full 2026-09-22; the full text was NOT read, so the cell systems, the specific G9a inhibitor used, the assay that scored 'rearrangement', and the size of each effect are not characterised here and must be read from the paper before any of this is built on. The readout is locus amplification and rearrangement in cell and mouse systems, not incidence of leukaemia and not an infant cohort - so this is a mechanism for how the initiating lesion arises, not evidence about any patient. It is also a therapy-induced-leukaemia model (doxorubicin), and this graph's index disease is de novo infant t(4;11) with an in-utero origin, where no anthracycline exposure precedes the break.

   ▸Evidence that argues against this (1)

   • The direction of this result is opposite to the downstream one and the two must not be merged into a single 'G9a is bad' claim. Here MORE H3K9me1/2 at the locus goes with MORE rearrangement and G9a INHIBITION suppresses the events; in established AML, G9a LOSS delays disease and forces differentiation. A single agent cannot be given both a prophylactic and a therapeutic rationale from these two papers, because they act at different times on different cells - one before the fusion exists, one after - and this graph's own `mech-kmt2a-loss-pattern-restraint-or-not` exists precisely to keep initiation and maintenance arguments apart. Lehnertz B et al. 2014
3. E3strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.paper

   G9a/EHMT2 is a selective regulator of fast-proliferating myeloid progenitors with no discernible function in haematopoietic stem cells: G9a-null mice are born at normal frequency with no overt haematological abnormality, while G9a deletion in secondary recipients of Hoxa9/Meis1 AML dropped circulating leukaemic cells by day 12 and raised median survival from 23.5 to 44 days, G9a deficiency similarly delayed MN1-induced and MLL-AF9-induced AML, ectopic wild-type G9a accelerated AML in vivo, the G9a/GLP inhibitor UNC0638 inhibited growth of G9a-heterozygous but not G9a-null AML cells, both G9a deletion and UNC0638 drove myeloid differentiation, and primary human AML cells were sensitive to G9a inhibition.

   The methyltransferase G9a regulates HoxA9-dependent transcription in AML. · Lehnertz B, Pabst C, Su L, Miller M, Liu F, Yi L, Zhang R, Krosl J, Yung E, Kirschner J, Rosten P, Underhill TM, Jin J, Hebert J, Sauvageau G, Humphries RK, Rossi FM · Genes and Development 2014
   PMID 24532712DOI
   ▸What this evidence cannot show

   The primary system is MURINE MYELOID leukaemia driven by Hoxa9/Meis1 co-transduction (A9M), not a KMT2A fusion and not B-lineage. The MLL-AF9 arm exists but is in Supplemental Fig. S5a,b - one supplementary panel, described in the text as 'similarly delayed', with no separate effect size given in the main text - so 'shown in MLL-AF9' here means one supplementary survival experiment, not a characterised result. There is no MLL-AF4 or t(4;11) arm and no B-lineage arm anywhere in the paper. Full text read via Europe PMC PMC3937511 on 2026-09-22 for these details.

   ▸Evidence that argues against this (1)

   • This graph's own H3K9 entry argues the opposite therapeutic direction and names the same compound. `mark-h3k9me3` records that H3K9 methylation RESTRAINS the MLL-fusion programme - SETDB1 overexpression in mouse MLL-AF9 cells silences a gene set whose promoters are H3K9-hypomethylated in AML patient samples - and that UNC0638 lowers H3K9me2 AND H3K9me3 together, so it concludes that a G9a or SETDB1 inhibitor is a candidate for making a KMT2A-rearranged leukaemia worse. The two positions concern different marks (me1/2 versus me3) written by different enzymes, but they cannot both be tested with UNC0638, and any UNC0638 result in this disease is therefore directionally ambiguous. Oncofusion Knowledge Graph 2026
4. E4strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database

   EHMT2 is NOT essential in DepMap CRISPR (Chronos): mean gene effect -0.2260 across 1,258 screened lines with 7.5% of lines depleted below -0.5, and a lymphoid-tissue mean of -0.2563 over 97 lines; its heterodimer partner EHMT1 is likewise not a dependency in this project's KMT2A scan (mean -0.129 across all lines, rank 1,189).

   DepMap CRISPR (Chronos) gene effect served through the Open Targets Platform GraphQL API, field depMapEssentiality, fetched 2026-09-22 with approvedSymbol read back as EHMT2; EHMT1 and CDYL values joined from analyses/depmap-kmt2a-selective-scan-full/results.json · Broad Institute DepMap via Open Targets Platform 2026
   ▸What this evidence cannot show

   Rule 83: the API does not state the DepMap release, so only the fetch date is on record, and 1,258 is EHMT2's own screen count. In this project's own KMT2A scan the KMT2A-rearranged arm is -0.277 against -0.268 in other leukaemia lines - a difference of -0.009 Chronos units at rank 7,974 of 17,728, which is noise and is recorded only so that the rank is never quoted as selectivity. A pooled viability screen is also the wrong assay for an enzyme whose published phenotype is a differentiation shift rather than death, and EHMT1/GLP is an obligate heterodimer partner, so single-gene knockouts of either may be buffered by the other.

   ▸Evidence that argues against this (1)

   • The genetic experiments say the opposite of the screen, and they say it in the system that matters. G9a deletion in established Hoxa9/Meis1 AML nearly doubled median survival (44 d against 23.5 d) and forced myeloid differentiation, and ectopic wild-type G9a ACCELERATED disease - none of which a pan-cancer pooled dropout screen scoring -0.23 would predict. Where a 1,258-line average and a lineage-matched conditional knockout disagree, the knockout is the measurement about this disease and the average is a prior. Lehnertz B et al. 2014
5. E5strong confidence 4 of 5: Replicated in peer-reviewed work by more than one group.database

   The G9a experiment in an MLL fusion is unreachable by any MLL or KMT2A query: EHMT2[tiab] = 401 and G9a[tiab] = 1,190, EHMT2[tiab] AND KMT2A[tiab] = 3 (37788669 plus two unrelated records, a mangrove-rivulus Kdm/Kmt gene-family survey and a nucleoporin imprinting paper), G9a[tiab] AND MLL[tiab] = 9 and EHMT2[tiab] AND MLL[tiab] = 3 - and PMID 24532712, whose supplementary figure carries the MLL-AF9 survival experiment, appears in NONE of those lists; it is reachable as EHMT2[tiab] AND HOXA9[tiab] = 1.

   NCBI E-utilities esearch, db=pubmed, counts and id lists with querytranslation and warninglist read back, run 2026-09-22; load-bearing zeros run twice · NCBI PubMed 2026
   ▸What this evidence cannot show

   Titles and abstracts only. The two query tokens are not interchangeable and both are needed: G9a[tiab] returns 1,190 records against EHMT2[tiab] at 401, and the MLL intersection is 9 for the alias against 3 for the approved symbol, so here the RETIRED/common name is the better operand - the same shape as registry row 122 (ELP1/IKBKAP), where the approved symbol won on raw count but 40.6% of the literature was reachable only through the retired one. Compare id sets, not counts: the EHMT2 AND KMT2A set of 3 is 2/3 off-topic.

   ▸Evidence that argues against this (1)

   • Widening the name list does not fix this one, which is the point of rule 80 rather than rule 71. The MLL-AF9 result for G9a is in Supplemental Fig. S5 and is mentioned in the body text as one clause; no synonym of MLL, KMT2A or any fusion name appears in that paper's title or abstract, so no abstract-level query of any spelling can reach it. Only opening the paper, or a full-text index, finds it - and the beat that reached it here did so through HOXA9, a gene name that has nothing to do with the query it was answering. Lehnertz B et al. 2014
6. E6moderate confidence 3 of 5: One strong study, or several smaller consistent ones.database

   The REST-to-G9a silencing route that this graph found from the Mediator end has a third required component: EHMT2 interacts with REST ONLY in the presence of CDYL, and a complex exists containing at least CDYL, REST, WIZ, SETDB1, EHMT1 and EHMT2, with transcriptional repression by REST-CDYL via recruitment of EHMT2 described as potentially important in transformation suppression.

   UniProtKB Q96KQ7 and Q13127 SUBUNIT and FUNCTION comments - 'Interacts with REST only in the presence of CDYL (PubMed:19061646)' and 'Part of a complex containing at least CDYL, REST, WIZ, SETB1, EHMT1 and EHMT2 (PubMed:19061646)', and REST's own FUNCTION note 'Transcriptional repression by REST-CDYL via the recruitment of histone methyltransferase EHMT2 may be important in transformation suppression', read 2026-09-22 · UniProt Consortium 2026
   ▸What this evidence cannot show

   This is a UniProt curation of PMID 19061646, and that paper was NOT retrieved for this entry, so the CDYL requirement is recorded at database strength only - rule 91 applies and the primary paper should be read before the condition is built on. Note also that UniProt spells the fifth member 'SETB1', which is almost certainly SETDB1; that reading is an inference from context and not a database value. The system in the underlying literature is neuronal-gene silencing in non-neuronal cells, not leukaemia, and the KMT2A-fusion relevance of this complex is zero at present.

   ▸Evidence that argues against this (1)

   • The Mediator-subunit route this graph records may not be the one that carries G9a. `gene-med19` itself notes that Mediator's contribution to REST silencing has also been placed on MED12 rather than on MED19/MED26, in work from the same laboratory - so which subunit carries the G9a route is contested inside one group's own output, and the CDYL requirement adds a second component that neither Mediator paper includes. A route with a contested adaptor and an undocumented cofactor requirement is not yet a route. Oncofusion Knowledge Graph 2026

• ← related toASB7 (Q9H672) - the brake on H3K9me3's own positive feedback loop, released by CDK1 in mitosis, and it reached this graph as the flattest dependency in the read queue

  G9a/EHMT2 is the H3K9 methyltransferase this graph's brake mechanism uses; ASB7 controls a DIFFERENT H3K9 writer, SUV39H1, by degradation rather than by catalysis

• related to →H3K9me3 - measured in KMT2A::AFF1 cells, in a supplementary figure, and it rises GENERALLY where H3K27ac falls SPECIFICALLY in the same experiment

  G9a writes H3K9me1/me2 and NOT me3; the two marks point opposite ways in MLL-fusion AML

• regulates →KMT2A / MLL1 — the gene that breaks

  H3K9me1/2 balance at the KMT2A LOCUS controls its amplification and rearrangement

• related to →KMT2A major breakpoint cluster region (BCR1): intron 7 - exon 13, ~8.3 kb

  a chromatin-mark layer on the rearrangement mechanism this graph models by TOP2 and replication stress

• related to →Doxorubicin - named in twenty entries of this graph, cuts the KMT2A breakpoint where etoposide does, and appears nowhere in Interfant-06

  doxorubicin suppresses KDM3B and CTCF; G9a inhibition suppresses the KMT2A events it causes

• binds →HOXA9 — required for survival of established MLL-rearranged leukemia, and dispensable for starting it

  G9a interacts with HoxA9 and supports HoxA9-dependent transcription in AML

• related to →MEIS1 — rate-limiting for MLL leukemia stem cells, a surviving co-dependency, and embryonic-lethal when removed

  the primary model is Hoxa9/Meis1 (A9M) murine AML

• related to →KMT2A-rearranged acute myeloid leukaemia in children and infants

  G9a deficiency delayed MLL-AF9-induced AML (supplementary figure only)

• related to →MED19 - the Mediator subunit curated against an engineered MLL-AFF4 fusion protein, and the one REST uses to deliver G9a

  MED19 and MED26 are the Mediator subunits REST uses to deliver G9a; the route also requires CDYL

• related to →MED26 — a common-essential Mediator subunit with a solved 87-residue domain and no ChEMBL target entity at all

  the second Mediator subunit in the REST-to-G9a route

• regulates →TP53 - named in 84 entries of this graph, absent from the index genome study of the disease, and four different things depending on when you look

  G9a dimethylates p53 at Lys-373 - a non-histone substrate

• binds →HDAC1 - a class I eraser that has no dependency claim of its own in this disease, because removing it alone does nothing until HDAC2 goes too

  TRIM28/HDAC1/HDAC2/EHMT2 corepressor complex; HDAC1 is also a G9a substrate

• binds →HDAC2 - the buffering half of the class I pair, and the one class I enzyme a drug result in acute leukaemia has actually been pinned to, in AML rather than in a KMT2A fusion

  TRIM28/HDAC1/HDAC2/EHMT2 corepressor complex

• expressed in →Fetal liver haematopoietic stem and progenitor cell — the leading candidate cell of origin for t(4;11) infant ALL

  UniProt records HIGH expression in FETAL LIVER, with a LOWER level in bone marrow

• related to →Chromocentres and pericentromeric satellite heterochromatin — no KMT2A evidence, but the AT-hook argument says there should be

  G9a writes the EUCHROMATIC H3K9 methylation, explicitly in contrast to pericentromeric heterochromatin

• related to →RNF2 (RING1B) — the one catalytic subunit shared by every PRC1 flavour, which is why the genetics point two ways at once

  both are members of the G0-phase E2F6.com-1 complex

• related to →Plain-English: chromatin and epigenetics — the sticky notes on the DNA, and which ones matter here

  a writer that also reads its own mark

• ← recruitsREST / NRSF - the transcription factor that delivers G9a to chromatin through MED19 and MED26, is a paediatric cancer gene, and is the worst search token in this corpus

  REST recruits EHMT1/EHMT2 for H3K9me2 silencing - but ONLY in the presence of CDYL

• regulates →H3K9me1 - the only H3K9 state whose level at the KMT2A breakpoint cluster region has ever been measured, and it is LOW there until KDM3B is removed

  G9a/EHMT2 writes H3K9me1 in euchromatin; its ANK repeats also read it

• regulates →H3K9me2 - named sixty times in this graph with no entry of its own, the mark that defines the lamina-associated domains the fusion's targets sit beside, and the only H3K9 methylation state ever measured at a KMT2A fusion's own binding sites

  G9a/EHMT2-EHMT1 heterodimer writes H3K9me2 in euchromatin

• ← related toKDM3B - the H3K9me1/2 demethylase that sits on the KMT2A breakpoint cluster region and keeps it from rearranging, and the only leukaemia gene in this graph whose direction of effect flips between the myeloid and the lymphoid literature

  opposed arms of one H3K9me1/2 balance; G9a depletion or dual G9a/GLP inhibition rescues the KDM3B-loss phenotype, EHMT1 depletion does not

• part of →The H3K9me1/2-CTCF brake: a KDM3B-G9a balance keeps the KMT2A breakpoint cluster region methyl-poor and CTCF-bound, and losing either end makes the locus amplify and rearrange - in epithelium, in primary cord-blood progenitors and in mice

  the writer arm: G9a depletion or dual G9a/GLP inhibition prevents the event, G9a over-expression alone causes it

• ← inhibitsUNC0642 - the G9a/GLP chemical probe that abolished doxorubicin-induced Kmt2a copy gain in mice, and the one intervention in this beat that has been given to an animal

  first in vivo G9a/GLP chemical probe; dual EHMT2 and EHMT1 inhibitor

• ← testsRUN: the eight unreported H3K9me3 samples of GSE210480, read at the KMT2A CTCF element - H3K9me3 does not move where H3K9me1 moves 3.9-fold, and the me1/me2 positive controls passed

  siG9A arm, two independent siRNAs; no H3K9me3 effect at the element (0.98x)

• ← related toAt the KMT2A anchor itself, KDM3B loss does not raise H3K9me1 AND H3K9me2 - it raises me1 3.9-fold while me2 FALLS to 0.44x, an inversion of the mono-to-di ratio that the paper's abstract describes as a joint increase

  G9a writes me1 and me2 processively, so mono accumulating without di implies a blocked second round at this element

• ← related toThe H3K9me3 null that appears to exonerate prophylactic G9a inhibition was measured in retinal pigment epithelium, at 1 kb resolution, with n = 2 - and the warning it appears to lift lives in a leukaemic blast

  the paper's genetics point at EHMT2 and not EHMT1, so UNC0642's GLP half is added exposure with no mechanistic role

• ← testsRUN: ten ENCODE peak sets and fourteen signal tracks read at the KMT2A locus in two blood cells - not one called peak for SETDB1, CBX5, CBX1, CBX3 or TRIM28 anywhere in the gene, while G9a/EHMT2 and TRIM28 both sit on the KMT2A PROMOTER and neither touches the breakpoint cluster region

  THE POSITIVE RESULT: G9a/EHMT2 has four peaks clustered at the KMT2A promoter and one 20 kb upstream in K562, and none at the breakpoint cluster region or the exon-11 CTCF element - 45 to 53 kb away

• ← related toThe off-target surface, measured rather than feared: KDM3B depletion moves CTCF occupancy at 84.6% of 12,443 RPE anchors and halves it at 1,436 - so an H3K9-axis drug given to prevent one rearrangement is remodelling loop architecture genome-wide, and preferentially at the strongest anchors

  G9a is the arm actually proposed as a drug target and its own CTCF footprint has never been measured in any cell

• ← testsThe deposit this graph filed as 'four SUV39H1 ChIP-seqs' is actually 77 samples in MCF-7 carrying the ENTIRE H3K9 axis in one cell type and one pipeline - writer SUV39H1, writer G9a/EHMT2, reader HP1-alpha/CBX5 (wild-type AND two phospho-mutants), co-repressor KAP1/TRIM28, erasers KDM4B and KDM4C, the mark H3K9me3, and an shKDM4B+shKDM4C eraser-knockdown arm

  G9a occupancy in MCF-7, same cells and pipeline as SUV39H1 and H3K9me3

• ← related toThe H3K9me3 this graph argues about at the KMT2A exon-11 anchor is not deposited by SUV39H1: in SU-DIPG13 the mark sits at the 57.8th percentile of class-matched CTCF anchors while the writer reads an EXPLICIT ZERO in the same nuclei, and genome-wide the two barely track each other (Spearman 0.148, 0.078)

  the second alternative writer, never read at this element - four deposited MCF-7 samples would test it

• ← related toRUN: G9a/EHMT2 reads the 95.8th percentile of 56,171 same-cell-type CTCF anchors at the KMT2A promoter in all four MCF-7 ChIP-seq tracks, and a median 33.42nd at the exon-11 anchor - so the cheapest test of hyp-q114's third prediction REFUTES it

  first signal-level occupancy of G9a/EHMT2 at the KMT2A locus in any cell type: 95.8th percentile at the promoter, ordinary at the anchor

• ← related toTHE POSITIVE RESULT: the writer G9a, the reader HP1a and the scaffold KAP1 all sit on the KMT2A PROMOTER in MCF-7 at the 87.6th-98.9th percentile of 2,000 ACTIVE promoters - surviving the control that says an active promoter is ChIP-hot for everything - while none of the three is above the 75th percentile at the exon-11 anchor 52 kb away

  G9a at the KMT2A promoter, 87.6th-88.7th percentile of 2,000 active promoters, 1.34x-1.44x the median active promoter

• ← related toGiving a G9a/GLP inhibitor alongside etoposide or doxorubicin removes a DNA-repair licence at the moment TOP2 breaks are being made - and the sibling compound UNC-0638 is measured to block BRCA1/BARD1 arrival at a break, in children on curative therapy

  G9a/EHMT2 is the target whose loss impairs HR by three independent published routes