Corpus 1,272 assessed · 1,173 scored · 643 reproduced ≥75 · 168 flagged ·∅ 74.1/100
← New search

Dynamics and regulation of mitotic chromatin accessibility bookmarking at single-cell resolution.

Sci Adv · 2023
L2 50/100 3/4
Why this verdict

The main result did not reproduce in this reproduction attempt. Where our recomputation produced values that differ from the published ones, those discrepancies are listed below. This is a single automated attempt — not peer review and not a finding of error or misconduct — and differences can also arise from data access, undocumented parameters or the computing environment. The verdict can be contested via “report an error”.

Reproduced on the brainbox compute brainarbeit.com
Scoring basis — itemised

Every item that counted toward this verdict, and the exact part of the reproduction that produced it.

Main result did not reproduce
Decisive
From: Q8 · Severity of the miss (overall human judgment) 🔴
✓ What held up
  • Nothing in this column.
What did not (or only partly)
  • 🔴Could not use the authors’ exact input data
  • 🔴Reported values were only indirectly comparable
  • 🟡A deviation arose in the data or preprocessing
  • 🔴A deviation was attributed to the published material
  • 🟡Reported values were not (fully) derivable from the shared data
  • 🔴The deviation was non-trivial in magnitude
  • 🟡The central claim did not (fully) hold under reproduction
  • 🔴Overall, the reproduction showed a material discrepancy
How its reproducibility compares
50/100
Reproducibility score
1.4 SD below mean
vs. all fields · 1173 studies
🎯 Scores higher than 8% of all assessed papers rank 1026 of 1173 scored

A 0–100 reproducibility-quality score from the per-question grades, shown as a z-score: standard deviations above (+) or below (−) the mean of comparable assessments.

Reproduction agent’s raw note

PARTIAL reproduction. The paper's OWN computational results (C2-C7: scATAC cell/peak counts, bookmarking fraction, NF-YA overlap, R=0.94/0.70 correlations) are NOT reproducible from shipped artifacts: real code is github.com/QuKunLab/Mitosis (commit a8f91c6) = 5 plotting-only notebooks reading a .«path» dir of ~40 processed intermediate files that are deposited NOWHERE (not repo/Zenodo/GSA/supplementary), with no upstream pipeline scripts -> docs_insufficient for those. Real raw data is GSA CRA003844 (NOT the registry's GSE92846, a mis-harvest) = 131GB raw PE FASTQ only. Per the brief's third-party-tool rule (P16), ONE clearly-specified result was reproduced end-to-end on the paper's own raw data: NF-YA binding-site count (reported 5088) via fastp->bowtie2(end-to-end,very-sensitive,GRCh38_noalt_as)->samtools filter(MAPQ30,proper-pair,dedup)->MACS2 callpeak(BAMPE,q0.05) vs IgG, on runs CRR609413/CRR609414 vs CRR609415. The well-powered replicate NFYAR2 (33.1M pairs) gives 5188 peaks -> within ~2% of the reported 5088; the shallower NFYAR1 gives 2793; reproducible overlap 1845. Graded PARTIAL (not exact) because genome build, MACS version/thresholds, and single-rep-vs-pooled/IDR definition are unspecified in the paper, and the two replicates differ ~1.9x by depth -- but one replicate reproduces the headline number near-exactly, which is strong positive evidence and argues against fabrication of that value. Tooling: bowtie2 2.5.5, samtools 1.21, MACS2 2.2.9.1 (needed a __*_finite LD_PRELOAD math shim for glibc compat), fastp 1.1.0, bedtools 2.31.1. Compute on «our HPC» COMPUTE nodes («job» align + 2225104 peaks; 2220320/2221952 were failed earlier attempts, fixed); all data on «infra». NOT attempted: C2-C7 (out of scope, undocumented + absent inputs). No fabrication asserted.

💻 Code ↗ 🗄 Data: GSE92846

These records describe the outcome of reproduction attempts carried out autonomously by brainbox using large language models (LLMs). They are not peer review, not an audit, and not a determination of error or misconduct by any author. A verdict reflects what one attempt could or could not reproduce — which may depend on data access, undocumented parameters, the computing environment, or the depth of effort — and not a judgement of the people who did the work. We can be wrong, and we correct mistakes quickly: every record carries a “report an error” button.

Assessment versions

Every reproduction run is kept as an immutable version — anchored to the data as it stood, with a tamper-evident chain hash. A rerun (e.g. after an author updates a deposit) adds a new version; the previous one stays on record.

  1. v1 current initial assessment
    assessed: 2026-06-15 ⛓ 3a3bad389a91
✎ I am an author of this paper

Updated or fixed a deposit, or is there an erratum? Ask us to re-run the metrics. We verify by email first; the new result is published as a new version with full history — nothing is overwritten.

Reason for the rerun

We email you a confirmation link first. The rerun is an objective re-measurement — it cannot change the verdict in your favour, only ask us to look again.

Provenance — full disclosure

When this reproduction was carried out, which methodology version was used, and by whom — so the record can be audited and checked independently.

Reproduced
2026-06-23
Rubric version
v1.0
Assessed by
🤖 AI curator · claude (ai-curator room) · v1.0 · run #1 2026-06-15
👤 1 human curator(s) · Level L2 2026-06-15
Last updated
2026-08-05

Provisional, curator- or AI-assessed, and independently checkable. A reproduction outcome states what one attempt could reproduce — not a judgement of the authors.

Deep full-text extraction

Model: sonnet
Founding hypothesis

The paper investigates whether and how chromatin accessibility is dynamically retained ('bookmarked') at specific genomic regions throughout mitosis, and what regulates the reestablishment of transcription after mitotic exit.

Core claims
  • Chromatin accessibility continually decreases from mitotic entry until metaphase, then gradually increases as chromosomes segregate. finding
  • A subset of chromatin regions (~7%, n=2249) remain accessible throughout all of mitosis, defined as 'bookmarked regions'. finding
  • Bookmarked regions are enriched near promoters/TSS and are associated with genes that reactivate rapidly (first-wave genes) after mitotic exit. finding
  • NF-YA preferentially occupies bookmarked regions and functions as a mitotic bookmarking transcription factor contributing to post-mitotic transcriptional reactivation. mechanism
  • Pseudotime-aligned single-cell ATAC-seq (scATAC-seq) reveals mitotic bookmarking dynamics that bulk ATAC-seq cannot detect. method
  • TFs split into mitotically lost (e.g., CTCF, PAX5, ASCL1) and mitotically enriched (e.g., RUNX2, MYC, NF-YA) groups based on correlation of motif enrichment with pseudotime. finding
  • Opening rate of chromatin regions positively correlates with proportion and expression of first-wave reactivated genes. finding
Experimental setups
Assay System Perturbation Readout Platform
scATAC-seq L02 human liver cells (unsynchronized H3pS10+ and RO3306/MG132/blebbistatin-synchronized) drug synchronization (RO3306, MG132, blebbistatin) / FACS sorting chromatin accessibility peaks across pseudotime APEC algorithm, Slingshot pseudotime inference
Immunofluorescence imaging L02 cells (FACS-sorted H3pS10+ population) none mitotic phase purity/staging
ATAC-see (imaging-based ATAC) L02, HepG2, HUH7 cells none (staged by mitotic phase) chromatin accessibility signal quantified vs DAPI
Bulk ATAC-seq L02 cells nocodazole arrest vs unsynchronized chromatin accessibility peaks
Motif enrichment / TF occupancy analysis L02 scATAC-seq peaks (JASPAR database) none (bioinformatic) TF enrichment score correlated with pseudotime hypergeometric test-based motif scanning
Public ATAC-seq dataset comparison 135 tissues/cell lines (ENCODE) none overlap with bookmarked regions ENCODE
Pol II ChIP-seq dataset comparison 34 tissues/cell lines (ENCODE) none overlap between bookmarked regions and Pol II binding sites ENCODE
EU-RNA-seq (nascent transcript) reanalysis human hepatoma cells and U2OS osteosarcoma cells (public datasets) mitotic block release (timepoints 0-300 min) gene expression of first-/second-wave reactivated genes
Key results
  • Chromatin accessibility decreased by ~75% during the prophase-metaphase transition before increasing after metaphase ~75% decrease
  • ~7% of initially open chromatin regions (n=2249) remained accessible throughout mitosis, defined as bookmarked regions 7% (n=2249)
  • ~79% of bookmarked regions located in gene promoters near TSS 79%
  • ~94% of bookmarked regions were open in public ATAC-seq datasets across 135 tissues/cell lines 94%
  • Significant overlap between bookmarked regions and common Pol II binding sites from ChIP-seq of 34 tissues/cell lines P<0.0001
  • Gene expression near bookmarked regions was significantly higher than unbookmarked regions at 80 min after mitotic release, but not in interphase P<0.0001
  • Opening rate of chromatin bins positively correlated with proportion and expression of first-wave genes R=0.94 (both)
  • Identified 110 mitotically lost TFs (ρ<0) and 131 mitotically enriched TFs (ρ>0), including known bookmarking factors RUNX2 and MYC 110 vs 131 TFs
Key statistics
  • count 6538 (total mitotic L02 cells profiled by scATAC-seq)
  • count 30,671 (total accessible chromatin peaks identified)
  • count 2249 (number of bookmarked regions (~7% of open regions))
  • fold_change ~75% decrease (chromatin accessibility decline during prophase-metaphase transition)
  • correlation R = 0.94 (opening rate vs proportion of first-wave genes)
  • correlation R = 0.94 (opening rate vs expression of reactivated genes at 80 min post-release)
  • pvalue P < 0.0001 (overlap of bookmarked regions with Pol II ChIP-seq binding sites (chi-square test))
  • pvalue P < 0.0001 (expression difference of reactivated genes near bookmarked vs unbookmarked regions (Student's t test))

Statistical methods review

Model: sonnet

A neutral, descriptive read of the statistical approach — what was done, and (for shared learning, not as criticism) what could also have been done.

The study applied single-cell ATAC-seq (scATAC-seq) to 6538 mitotic L02 human liver cells, using the APEC algorithm for peak calling and the Slingshot algorithm for pseudotime trajectory inference to represent continuous mitotic progression from prophase to anaphase. Chromatin accessibility dynamics across 30,671 peaks were characterized along the pseudotime axis, and bookmarked versus unbookmarked regions were compared using chi-square tests and two-sided Student's t-tests. Transcription factor motif enrichment was scored via hypergeometric test-based motif scanning against the JASPAR database, and Pearson correlations were used to relate chromatin opening rates to gene reactivation dynamics.

Replicationunclear Sample size6538 total cells: 5448 unsynchronized H3pS10+, 352 RO3306-treated prophase, 363 MG132-treated metaphase, 375 blebbistatin-treated anaphase; n=3 per group for ATAC-see immunofluorescence quantification; number of biological replicates for scATAC-seq experiment(s) not stated in the provided text GroupsMitotic stages (prophase, prometaphase, metaphase, anaphase) via pseudotime; bookmarked vs. unbookmarked chromatin regions; gene sets near bookmarked vs. unbookmarked vs. bulk-detected regions; post-mitotic release time points vs. interphase Pairingunpaired Randomization/blindingnot stated DispersionSD Exact p-valuesno Effect sizesno Confidence intervalsno Multiplicity correctionnone stated
Statistical tests used
Test Applied to n Assumptions
Chi-square test Overlap between bookmarked regions and open sites from 181 ATAC-seq datasets (Fig. 2C) and overlap with common Pol II ChIP-seq binding sites from 34 tissues/cell lines (Fig. 2D) not stated
Two-sided Student's t-test Z-scaled expression of reactivated genes near bookmarked vs. unbookmarked vs. bulk-detected regions at 80 min after mitotic release (Fig. 2F) and in interphase cells (Fig. 2G) not stated
Pearson correlation (R) Opening rate of 13 chromatin bins vs. proportion of first-wave genes in associated genes (Fig. 2I) and vs. expression of reactivated genes at 80 min post-release (Fig. 2J) 13 bins not stated
Pearson correlation coefficient (ρ) TF motif enrichment score vs. pseudotime for each of 241 TFs to classify mitotically lost (ρ < 0) vs. mitotically enriched (ρ > 0) TFs (Fig. 3A) 241 TFs not stated
Pearson correlation TF enrichment time vs. proportion of bookmarked regions in TF-targeted regions across mitotically enriched TFs (Fig. 3C) not stated
Hypergeometric test-based motif scanning Inferring per-cell TF enrichment scores from JASPAR motifs across scATAC-seq peaks throughout the pseudotime trajectory not stated
Approaches that could also have been used
  • Two-sided Student's t-test was used to compare Z-scaled expression values between gene sets associated with bookmarked vs. unbookmarked chromatin regions
    Could also: A Mann-Whitney U (Wilcoxon rank-sum) test could also have been applied — RNA-seq-derived expression values, even when Z-scaled, often have skewed distributions; a non-parametric rank-based test makes no normality assumption and is robust to outliers, which is a widely used alternative for comparing gene expression distributions in genomics
  • Pearson correlation was used to relate chromatin bin opening rates to gene reactivation metrics across 13 bins
    Could also: Spearman rank correlation could also have been used — With only 13 data points and a potentially monotonic but not strictly linear relationship, Spearman correlation does not assume linearity or homoscedasticity, and provides a complementary measure of association that is also commonly reported alongside Pearson R in genomic studies
  • Chi-square tests were used to assess statistical significance of overlaps between bookmarked regions and large ENCODE ATAC-seq or ChIP-seq datasets
    Could also: Fisher's exact test, or a permutation-based overlap test (e.g., as implemented in BEDTools shuffle or LOLA), could also have been used — Fisher's exact test does not rely on large-sample chi-square approximations and is exact by construction; permutation-based tests additionally account for genomic features such as GC content and mappability that affect the null distribution of region overlaps
  • Multiple chi-square tests, t-tests, and Pearson correlations were performed across several comparisons without an explicitly stated multiple comparison correction
    Could also: A Benjamini-Hochberg FDR correction could also have been applied across the family of tests within each analysis — When several related hypothesis tests are conducted, controlling the false discovery rate is a widely used approach to quantify the expected proportion of false positives among significant findings, and is standard practice in high-dimensional genomics studies
  • Dispersion for ATAC-see signal quantification (n=3 per group across five mitotic phases) was reported as mean ± SD
    Could also: A 95% confidence interval could also have been reported — At n=3 per group, a 95% CI directly conveys the uncertainty around the estimated mean and is often considered more informative for inferential purposes than SD, which describes sample spread rather than estimation precision
  • Pseudotime trajectory inference was performed using the Slingshot algorithm in two-dimensional space
    Could also: Other trajectory inference methods such as Monocle 3 (with UMAP embedding) or PAGA could also have been applied — Different trajectory algorithms make distinct assumptions about manifold topology, branching structure, and dimensionality reduction; applying an alternative method is a standard sensitivity check to assess the robustness of the inferred pseudotime ordering and downstream chromatin accessibility dynamics
Software: APEC (accessibility pattern-based epigenomic clustering) · Slingshot · GREAT · HOMER

Result convergence & founder nodes

Findings this paper shares with others that ran a comparable experiment. A node’s strength is how many independent papers report it (replication breadth) — not how often it is cited, so a heavily-replicated but under-cited founder still stands out.

Citation network

Where this publication sits in the reproducibility-weighted citation graph — what it is built on, and what is built on it. Citation data from OpenAlex.

Citations
25
Impact: medium
Foundation confidence
None of its references are in our reproducibility record yet — its foundation cannot be assessed.
Topics

No assessed neighbours yet — the network grows as more papers are assessed.

Data lineage

The datasets this paper uses (text-mined from the full text via Europe PMC), and which other assessed papers stand on the same data. A shared dataset is a factual link — not a judgement.

C10329 ENA in Methods (http://purl.org/orb/Methods)
no other assessed paper uses this yet
C10330 ENA in Methods (http://purl.org/orb/Methods)
no other assessed paper uses this yet
C10337 ENA in Methods (http://purl.org/orb/Methods)
no other assessed paper uses this yet
E10345 ENA in Methods (http://purl.org/orb/Methods)
no other assessed paper uses this yet
GSE109962 GEO in Methods (http://purl.org/orb/Methods)
no other assessed paper uses this yet
GSE75066 GEO in Methods (http://purl.org/orb/Methods)
no other assessed paper uses this yet
GSE92846 GEO in Methods (http://purl.org/orb/Methods)
no other assessed paper uses this yet

Downstream reach in the literature

2 downstream papers · 3 datasets

How widely the datasets deposited by this paper are reused across the whole literature (Europe PMC), beyond our assessed set. This is a factual dependency map — reusing a public dataset is normal, good science. It is not a judgement on the downstream papers; the only verdict here is this paper's own, with its cited rationale.

GSE92846 GEO reused by 3 papers in the literature
Most-cited downstream papers:
GSE109962 GEO reused by 2 papers in the literature
Most-cited downstream papers:
GSE75066 GEO reused by 2 papers in the literature
Most-cited downstream papers:

What was reproduced

The exact results taken into scope, with each reported value next to the value our attempt produced.

Reproduction scope — pmid-36696508

Paper: Yu Q et al. "Dynamics and regulation of mitotic chromatin accessibility bookmarking at single-cell resolution." Sci Adv 2023;9:eadd2175. PMID 36696508 · PMCID PMC9876548 · DOI 10.1126/sciadv.add2175.

Provenance correction (important)

  • Registry/BRIEF listed data as geo:GSE92846 and code as github.com/QuKunLab/ATAC-pipe. Both are mis-harvested:
    • The paper's actual data is in GSA (CNCB/NGDC) accession CRA003844 (BioProject PRJCA004401) — not GEO. GSE92846 is an unrelated older GEO series; ATAC-pipe is only one tool the paper cites, not its repo.
    • The paper's actual analysis code repo is github.com/QuKunLab/Mitosis (pinned commit a8f91c62da7bbd476fe59113fcc0925a5b4ffd1e, v1.0.0, 2022-11-11), mirrored on Zenodo 10.5281/zenodo.7313683 (code-only zip, 2.8 MB). This is a provenance/harvest error, not an author fabrication.

What the shipped code actually is

QuKunLab/Mitosis = 5 Jupyter notebooks (s003_Fig1..Fig5.ipynb), plotting-only. Every notebook reads from a local .«path» directory of ~40 processed intermediate files (peak BEDs, count CSVs, annotation tables, pseudotime/TF tables, e.g. genes_scored_by_TSS_peaks.csv, merge.bed, idr_NFYA_p12_igg_bam_peaks.bed, results.deseq.csv, …).

These .«path» files are shipped NOWHERE — not in the repo, not in the Zenodo snapshot, not in GSA (raw FASTQ only), and not as accessible supplementary. No upstream pipeline scripts / workflow / parameters are provided to regenerate them. The notebooks therefore cannot be executed as shipped.

In scope vs out of scope (pipeline-derived results)

Reported result Pipeline In scope? Why
6,538 mitotic scATAC cells; 30,671 peaks APEC v1.1.0.11 scATAC + MACS2, then bespoke QC Out needs full APEC reprocessing of thousands of per-cell FASTQ runs + bespoke cell QC; no scripts → hard >20%
~2,249 bookmarked regions (~7%) bespoke scATAC accessibility-dynamics calc Out depends on absent .«path» + undocumented definition
R=0.94 (opening rate vs first-wave genes); R=0.70 (NF-YA vs scATAC) notebook Fig1/Fig3 on absent .«path» Out inputs (genes_scored_by_TSS_peaks.csv, etc.) not shipped
Trajectory / Palantir / Slingshot scATAC pipeline Out inputs not shipped
DESeq2 RNA-seq DE (results.deseq.csv) STAR+HTSeq+DESeq2 Out needs full RNA reprocessing; no scripts
5,088 NF-YA binding sites (NF-YA CUT&Tag/ChIP, MACS2) bowtie2 + MACS2 (+IDR) IN (attempted) well-specified standard bulk pipeline; raw data available as a small GSA subset (NFYAR1/2 + IgG); third-party-tool-on-paper's-data per BRIEF rule P16

Attempted reproduction (the one clear, low-cost data point)

Third-party standard pipeline on the paper's own raw data:

  • Data: GSA CRA003844 runs NFYAR1=CRR609413 (26.5M PE), NFYAR2=CRR609414 (33.1M PE), IgG control iggForYA=CRR609415 (29.3M PE). 150 bp PE NovaSeq.
  • Pipeline: fastp → bowtie2 (--end-to-end --very-sensitive, GRCh38_noalt_as) → samtools filter (proper-pair, MAPQ≥30, chr1-22/X/Y, dedup) → MACS2 callpeak -f BAMPE -g hs -q 0.05 (NFYA rep vs IgG) → IDR(0.05) across reps.
  • Compared against: paper's reported 5,088 NF-YA binding sites.
  • Caveats: genome build unspecified in paper (assumed GRCh38); exact MACS2 q/IDR thresholds and read-filtering not fully specified → expect partial / same-order-of-magnitude agreement, not exact. «our HPC» job 2178083.

Bottom line

The paper's own pipeline outputs are not reproducible from shipped artifacts (plotting-only code + absent processed inputs + no upstream scripts) → shipped-code path = docs_insufficient. We instead reproduce one clearly specified result (NF-YA binding sites) via a standard third-party pipeline on the paper's raw data, as an honest partial check. We do *

Figures / tables: Fig 4Fig 1Fig 3
C1
Reported
5088 NF-YA binding sites (Fig 4)
Reproduced
NFYAR2(CRR609414)=5188 peaks (within ~2% of 5088); NFYAR1(CRR609413)=2793; reproducible-overlap(recip50%)=1845. bowtie2->samtools-filter->MACS2 BAMPE q0.05 vs IgG CRR609415, GRCh38_noalt_as.
partial
C2
Reported
6538 mitotic scATAC cells
Reproduced
not attempted (out of scope: full APEC scATAC reprocessing, no scripts, absent .«path»)
partial
C3
Reported
30671 accessible peaks
Reproduced
not attempted (out of scope)
partial
C4
Reported
~2249 bookmarked regions (~7%)
Reproduced
not attempted (out of scope)
partial
C5
Reported
61% NF-YA overlap with bookmarked regions
Reproduced
not attempted (out of scope)
partial
C6
Reported
R=0.94 opening-rate vs first-wave genes (Fig 1)
Reproduced
not attempted (out of scope)
partial
C7
Reported
R=0.70 NF-YA CUT&Tag vs scATAC (Fig 3)
Reproduced
not attempted (out of scope)
partial

Assessments & scoring basis

Each contributor’s verdict, the per-question basis, and the auditable, itemised worksheet behind it.

🤖 AI curator · claude (ai-curator room) · v1.0 L1 50/100

An automated assessment. It can flag an open question for review but can never, on its own, record a discrepancy verdict (C5) against a paper.

🔴1. Data identity
🔴2. Endpoint comparability
🟡3. Location of the main deviation
🔴4. Cause of the deviation
🟡5. Derivability / plausibility
🟡6. Severity of the deviation
🟡7. Core claim
🟡8. Severity of the miss (overall human judgment)
👤 Schlein Lab (curation team) L2 19/100
🔴1. Data identity
🔴2. Endpoint comparability
🟡3. Location of the main deviation
🔴4. Cause of the deviation
🟡5. Derivability / plausibility
🔴6. Severity of the deviation
🟡7. Core claim
🔴8. Severity of the miss (overall human judgment)
Scoring basis — itemised

Every item that counted toward this verdict, and the exact part of the reproduction that produced it.

Main result did not reproduce
Decisive
From: Q8 · Severity of the miss (overall human judgment) 🔴
🤝
Reproduced automatically — and fairly

Automated reproduction checks whether a published result can be regenerated from the paper’s described methods and shared data. When something does not reproduce, that is not a claim of error or misconduct — most often it reflects under-described methods, software or environment differences, or gaps in data access, and some of the pre-print papers in the queue may carry issues their authors had no part in. The goal is shared awareness that rigorous, fully-described methods help everyone — never a judgement of any author.

Are you an author? We would genuinely like to hear from you — to clarify the record, add data or code, re-run the pipeline after an accession update, and publish your response right next to the assessment. Everything here is open and auditable.

🚩 Report an error in this record

Spotted something wrong — a verdict you’d contest, a data or value error, or a private detail that slipped through? Tell us, with a short justification. Authors and readers are equally welcome to write in; we review every report.

Prefer email, or the form below not working? Contact us at support@doesitreproduce.com.

Reproduction footprint

claude-opus-4-8

Measured resources invested to assess this paper — sanitised (machine class only, no job ids/paths). Compute = HPC accounting (SLURM); tokens = the AI agent's session.

334.1 k
tokens (I/O) · 23.3 M incl. cache
326 min
runtime · 23.12 CPU-h
11.2 GB
peak RAM
4 (3 failed)
HPC jobs
hummel
machine