H4K20me1 - the transcription- and replication-linked H4 state, whose gene-body level tracks Pol II; it HAS been profiled genome-wide in human haematopoietic stem cells twice, and a query scoped to 'leukemia' cannot see either dataset

Read at abstract level; the PMCID (PMC12219854) was confirmed and the full text was not opened this session, so the cell systems, the measured selectivity margin over H4K20me2 and the magnitude of the MCM-loading defect are not quoted. No haematopoietic or leukaemia system is named in the abstract. The finding that BAHCC1 depletion REDUCES H4K20me1 means reader and mark cannot be separated by depletion alone.

• BAHCC1's better-known reader module in this graph is the BAH domain reading H3K27me3, on which gene-bahcc1 rests. A protein with two histone-reader modules of opposite valence - repressive H3K27me3 and transcription/replication-linked H4K20me1 - cannot be assigned a single chromatin role, and any occupancy result for BAHCC1 is ambiguous between them unless both marks are profiled. Li D et al. 2025

Abstract level. The abstract gives 'H4K20me1/2' without a stated preference between the two states, so this entry does not claim MBTD1 prefers me1; UniProt Q05BQ5's function comment likewise records both. Human cell lines, not haematopoietic primary cells. This is the same evidence item as gene-mbtd1#E3 and is repeated here so the mark node stands alone.

• [ADDED BY THE COORDINATOR DURING SALVAGE, 2026-09-23 cycle 82. Q121 was killed by the account session limit before it wrote this field. This is a reasoned methodological limitation derived from the claim and its citation type, NOT a literature-sourced rebuttal - no counter-citation was fetched, and none is asserted.] Reader specificity for me1 AND me2 through the same domain is the problem, not the finding: a reader that binds both states cannot discriminate them, so MBTD1 occupancy cannot be used as evidence about which H4K20 state is present at a locus. The entry should not treat MBTD1 as an H4K20me1-specific probe. STRONGER COUNTER: none needed - the claim limits itself.

A 2010 REVIEW, cited only as a map for the three-state consensus, and it is out of date on the central point it makes about erasers: PHF8 was published as an H4K20me1 demethylase in Nature in the same year (E5) and RSBN1/KDM9 followed in 2017. The author list (Balakrishnan L, Milavetz B) was fetched from esummary on 2026-09-23 AFTER a first draft of this entry carried a fabricated author list for this record - recorded because the fabrication was caught by re-fetching, not by review. mark-h4k20me2 cites this same review for the same purpose.

• The review's statement that no H4K20 demethylase is known was superseded within the year by PHF8, and later by RSBN1/KDM9. Anything in this graph resting on 'H4K20 methylation is irreversible' is wrong and should be corrected against E5 and gene-rsbn1. Liu W et al. 2010

TITLE ONLY - the abstract was not retrieved this session. The title is quoted from a fetched esummary record, not completed from memory. Note the internal tension the title itself carries: the enzyme that writes the state now assigned to TRANSCRIPTION was first described as associated with SILENT chromatin, which is a 2002 framing this entry records rather than endorses. No haematopoietic system. This is cited as the writer-identity primary source that mark-h4k20me2 explicitly flags as missing for its own writer list.

• [ADDED BY THE COORDINATOR DURING SALVAGE, 2026-09-23 cycle 82. Q121 was killed by the account session limit before it wrote this field. This is a reasoned methodological limitation derived from the claim and its citation type, NOT a literature-sourced rebuttal - no counter-citation was fetched, and none is asserted.] A 2002-era biochemical characterisation. 'Associated with silent chromatin' was a reasonable reading then and the field has since assigned H4K20me1 roles in transcription, replication licensing and mitosis that sit awkwardly with a purely silencing account. Quoting the silent-chromatin association without that history would misrepresent current understanding. STRONGER COUNTER: a modern review or a genome-wide study contradicting the silencing assignment. Not fetched.

TITLE ONLY, from a fetched esummary record; the abstract was not retrieved and no cell system, substrate specificity margin or mechanism is quoted. PHF8 is a multi-substrate JmjC enzyme also reported to remove H3K9me1/me2, so 'the H4K20me1 demethylase' would overstate it - and any experiment perturbing PHF8 perturbs at least two marks. `PHF8[tiab] AND H4K20me1[tiab]` returns 23 records, so this is a substantial literature of which one title was read.

• [ADDED BY THE COORDINATOR DURING SALVAGE, 2026-09-23 cycle 82. Q121 was killed by the account session limit before it wrote this field. This is a reasoned methodological limitation derived from the claim and its citation type, NOT a literature-sourced rebuttal - no counter-citation was fetched, and none is asserted.] Demethylase assignments are notoriously substrate-promiscuous in vitro, and a cell-cycle-progression phenotype is downstream enough that it does not establish which H4K20 state the enzyme acts on in cells. PHF8 also has documented activity on other substrates. STRONGER COUNTER: a PHF8 substrate-specificity study, or a separation-of- function mutant. Not fetched.

Abstract level only; the full text was not opened and no accession for the deposited data was looked up, so no GEO series is named here. HUMAN PRIMARY HSPCs driven to erythroid maturation - the right species and a haematopoietic cell, but a lineage this project does not study and a normal differentiation rather than a leukaemia. The correlations are genome-wide correlations, not perturbations: nothing here shows H4K20me1 causes anything. The author list was fetched from esummary on 2026-09-23 (Murphy K, Getman M, Lv X, Murphy Z, Abid D, Rahman N; last author Steiner L) and is reproduced abbreviated.

• The paper's own exception class refutes any simple reading: a small set of highly expressed genes including beta-globin LOSES H4K20me1 while gaining Pol II. So gene-body H4K20me1 is not a monotone reporter of output, and a low H4K20me1 reading at a hyperactive locus is consistent with both low elongation and active PHF8 demethylation. Any KMT2A-fusion experiment on this mark needs a PHF8 arm. Murphy K et al. 2025

Abstract level only. HUMAN primary HSCs/HPCs, 2009 ChIP-seq depth and antibody generation, one lineage (erythroid) and one differentiation endpoint. H4K20me1 appears here as one of five co-associated features of a gene class, not as the subject of the study, and no effect size is given. The author list is from the fetched esummary record. This is a CORRELATIVE chromatin-signature result and this project's registry warns that 'bivalent' is a polysemous token near the menin literature - here it is used in the chromatin sense the paper intends.

• [ADDED BY THE COORDINATOR DURING SALVAGE, 2026-09-23 cycle 82. Q121 was killed by the account session limit before it wrote this field. This is a reasoned methodological limitation derived from the claim and its citation type, NOT a literature-sourced rebuttal - no counter-citation was fetched, and none is asserted.] Genome-wide correlative profiling through differentiation. Bivalency and H4K20me1 co-occurrence at a locus is an association across a population of cells and cannot establish that the marks co-occur on the same nucleosome or the same allele - the single-cell and allele-resolution versions of this measurement are what would settle it and this project has recorded that limitation for other marks. STRONGER COUNTER: an allele- or nucleosome-resolved measurement. Does not appear to exist for this mark.

Abstracts only for all five; no full text was opened, so a genome-wide panel hidden in a supplementary figure of any of them would not have been seen - which is precisely the limitation that makes the hematopoietic-scoped query in E6/E7 the more informative one. Two of the three substantive studies share a first author (Zhou HR) and a locus (Wnt5a), so the effective independent sample is smaller than three. This item closes the open question posed in mark-h4k20me2 and its verdict is partly a correction of that entry's broader claim.

• [ADDED BY THE COORDINATOR DURING SALVAGE, 2026-09-23 cycle 82. Q121 was killed by the account session limit before it wrote this field. This is a reasoned methodological limitation derived from the claim and its citation type, NOT a literature-sourced rebuttal - no counter-citation was fetched, and none is asserted.] Five records is a number so small that the distinction between 'three substantive' and 'two peripheral' rests on one reader's judgement of each paper, made in one pass. And rule 80 applies: H4K20me1 measured in a supplementary western in a leukaemia paper whose abstract never names the mark is invisible to the query. The count bounds indexed abstracts, not the literature. STRONGER COUNTER: full-text search. Not run.

CARRIED FROM SIBLING ENTRIES, NOT INDEPENDENTLY VERIFIED BY THIS BEAT. None of the three PMIDs behind this statement was fetched or read this session; they are cited here as the route to the claim and not as this entry's own sourcing. A beat extending the me1 node should open PMID 34185806 before any claim rests on the direction of the RSBN1 reaction. The URL field carries a repository placeholder because the source is local files.

• [ADDED BY THE COORDINATOR DURING SALVAGE, 2026-09-23 cycle 82. Q121 was killed by the account session limit before it wrote this field. This is a reasoned methodological limitation derived from the claim and its citation type, NOT a literature-sourced rebuttal - no counter-citation was fetched, and none is asserted.] THIS ITEM CITES THIS REPOSITORY RATHER THAN PRIMARY LITERATURE, which is the weakest citation form in this corpus and is flagged rather than accepted: the RSBN1/KDM9 activity should be read from the DPY-21 and human studies directly before it is relied on. An eraser-of-higher-states being a producer-of-me1 is also a statement about enzymatic activity, not about net cellular levels, which depend on the writer's activity too. STRONGER COUNTER: the primary RSBN1 papers, re-read.

TITLE ONLY. Mouse oocytes; no haematopoietic or human relevance is claimed. Confidence is weak deliberately: a title is not a result, and this is recorded because it is the second independent hint (with E1's BAHCC1 depletion lowering H4K20me1) that H4K20me1 readers stabilise their own ligand - a pattern that would confound every reader-depletion experiment on this mark.

• [ADDED BY THE COORDINATOR DURING SALVAGE, 2026-09-23 cycle 82. Q121 was killed by the account session limit before it wrote this field. This is a reasoned methodological limitation derived from the claim and its citation type, NOT a literature-sourced rebuttal - no counter-citation was fetched, and none is asserted.] Mouse oocyte meiosis again - different species, a specialised and transcriptionally quiescent chromatin state, and a cell type with no haematopoietic relevance. That a reader also protects the mark there is a mechanistic possibility elsewhere, not evidence. STRONGER COUNTER: the same reader-protects-mark test in a somatic cell.

Titles and abstracts only; supporting figures not searched, and for a MARK rather than an entity that is the limitation that matters most. The writer's symbol was run as KMT5A and SETD8 but NOT as PR-Set7, SET8, SET07 or PR/SET07, so the writer-side name list is incomplete and the writer-side zeros are weaker than the mark-side ones. No archive (GEO, PRIDE, ProteomeXchange) query was run at all.

• [ADDED BY THE COORDINATOR DURING SALVAGE, 2026-09-23 cycle 82. Q121 was killed by the account session limit before it wrote this field. This is a reasoned methodological limitation derived from the claim and its citation type, NOT a literature-sourced rebuttal - no counter-citation was fetched, and none is asserted.] A zero against KMT2A and MLL in title/abstract is the project's most familiar trap. Rule 80: the mark could be measured in any KMT2A paper's supplementary data and this query would never see it. And this corpus has recorded that H3K9me1/2 was reachable only from one of its two names, with 91% invisible from the other - so a two-name query on a mark is a weak bound. The zero justifies writing this entry; it does not establish that nobody has measured H4K20me1 in a KMT2A context. STRONGER COUNTER: full-text and supplementary search, plus alias forms such as H4K20me and H4K20monomethyl.