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

Patient-Derived Meningioma Organoids: A Reliable Model for Studying Human Tumor Pathophysiology.

Cancers (Basel) · 2025
L1 50/100 3/4
Why this verdict

The main results reproduced, with only marginal, non-material deviations.

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.

Supporting (toward a concern)
Content-critical question only partially held
+2 pts
From: Q8 · Severity of the miss (overall human judgment) 🟡
Minor / cosmetic deviation
+1 pts
From: Q4 · Cause of the deviation 🟡
Minor / cosmetic deviation
+1 pts
From: Q2 · Endpoint comparability 🟡
Concordant (toward reproduced)
Code + data deposited & functional
-2 pts
From: Data & code availability Available & functional
Total score +2
✓ What held up
  • Same input data as the authors
  • No relevant deviation in data/preprocessing
  • Reported values are derivable from the shared data
  • Any deviation was negligible
  • The central claim held under reproduction
What did not (or only partly)
  • 🟡Reported values were only indirectly comparable
  • 🟡A deviation was attributed to the published material
  • 🟡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 honest ceiling: the paper reports all bioinformatic results as FIGURES ONLY -- no numeric r, %, variance fractions, or DEG counts in text -- so no value is 1:1 pinnable; this is figure-only reporting, not by itself a fabrication flag). Compute ran on «our HPC» compute node n110 (SLURM 2225186, ExitCode 0) from the paper's OWN shipped processed matrix GSE287174_FPKM.csv.gz (21,014 genes x 65 samples = 33 tumor + 32 organoid). C1 (Fig 4b): within-pair tumor<->organoid Pearson r mean 0.942 (0.92-0.96 per patient) CONFIRMS the paper's strong-similarity claim; honest nuance = the unpaired background r is also high (0.918), so all samples are globally similar and the within-vs-cross margin is modest (+0.024). C2 (Fig 4a): PCA reproduces every stated qualitative point -- low per-PC variability (PC1 20.6%), a clear PC1 shift between tumor and organoid groups (the 'right cultural shift on PC1'), and minimal PC2 separation. METADATA CORRECTIONS carried from prior run and re-verified: the study's own data is GSE287174 (harvested GSE183655 is only the CIBERSORTx scRNA reference), and rnaseqmut is genuinely the authors' tool (third-party tool on own data, valid per P16). NOT ATTEMPTED in this pass: C3/Fig 5a rnaseqmut concordance (heavy, raw-FASTQ stretch -- being attempted next as a bonus), CIBERSORTx deconvolution Fig 4d (registration-gated web tool), and all wet-lab results (79% culture success, IHC/IF, drug response) which are out of scope. The prior run's interactive-2FA VPN block is fully resolved (tunnel centrally up). Grades provisional; a human reviewer signs off.

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.

✎ 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 2026-06-23
no human curator yet
Last updated
2026-07-31

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 study aims to develop and validate a standardized protocol for establishing patient-derived meningioma organoids (MEN-Os) that faithfully replicate the histopathological, molecular, and cellular features of human meningiomas.

Core claims
  • A standardized, reproducible protocol can establish meningioma organoids (MEN-Os) from patient-resected tumor tissue without mechanical/enzymatic dissociation, using serum-free medium lacking growth factors or exogenous ECM. method
  • MEN-Os preserve the histopathological architecture of parent tumors, including whorl patterns and psammoma bodies. finding
  • MEN-Os retain meningioma-specific marker expression (SSTR2, progesterone receptor) similar to parental tumors. finding
  • MEN-Os maintain proliferation rates (KI67) comparable to parent tissue. finding
  • MEN-Os preserve immune microenvironment cells, including macrophages. finding
  • MEN-Os retain transcriptional signatures of parent tumors, including expression of meningioma-associated genes NF2, CDKN2A, and TP53, and preserve inter- and intra-tumoral heterogeneity. finding
  • Deconvolution of bulk RNA-seq shows MEN-Os retain diverse cell populations (tumor cells, stem cells, endothelial cells, macrophages, dendritic cells) reflecting parent tumor composition. finding
  • Transcriptomic mutation analysis (SNPs/indels) shows high concordance between MEN-Os and original tumors. finding
Experimental setups
Assay System Perturbation Readout Platform
organoid culture establishment patient-resected meningioma tissue (15 of 19 samples, 11 grade 1, 4 grade 2) none (ex vivo culture, ROCK inhibitor for first 48h) spherical organoid formation, sustained growth for ≥2 weeks orbital shaker incubator, 37°C/5% CO2/90% humidity
brightfield growth imaging MEN-Os none 2D projected area and circularity over 0, 2, 4 weeks ImageJ software 2.14
H&E and immunohistochemistry parent tumor tissue and matched MEN-Os (paraffin-embedded sections) none cytoarchitecture, whorl patterns, psammoma bodies, SSTR2/PR/KI67/macrophage marker expression Histology and Molecular Pathology Core, Emory NPRC
immunofluorescence staining OCT-embedded tumor tissue and MEN-Os (25 µm cryosections) none marker localization/expression
bulk RNA sequencing parent tumor tissue and corresponding MEN-Os (after 4 weeks in culture) none transcriptome-wide gene expression, PCA, differential expression Illumina NovaSeq6000; nf-core pipeline, STAR, Salmon, DESeq2
bulk RNA deconvolution bulk RNA-seq data from tumors/MEN-Os plus scRNA-seq reference from six meningiomas (GSE183655) none relative proportions of tumor, stem, endothelial, and immune cell populations CIBERSORTx; Seurat V4; SingleCellExperiment V1.28
transcriptomic mutation analysis aligned RNA-seq reads from parent tumors and MEN-Os none SNP/indel calling and concordance of mutant transcript fractions rnaseqmut pipeline
Key results
  • MEN-Os were successfully established in 79% of samples 15/19 (79%)
  • MEN-Os maintained stable growth and structural integrity in culture for up to four weeks with minimal size increase, consistent with slow meningioma growth
  • MEN-Os displayed characteristic whorl patterns and psammoma bodies matching parent tumor histology
  • Strong SSTR2 positivity with heterogeneous expression observed in MEN-Os
  • PR positivity in MEN-Os was similar to parent tumors with insignificant differences in PR-positive cell counts
  • KI67 proliferation rate largely maintained between parent tissue and MEN-Os
  • PCA showed low PC1/PC2 variability and relative proximity between parent tumors and MEN-Os, with MEN-Os preserving individual tumor-specific expression patterns (e.g., NF2 expression clusters)
  • Deconvolution revealed retention of tumor, stem, endothelial, and immune (macrophage, dendritic) cell populations in MEN-Os at proportions reflecting parent tumors
Key statistics
  • count 39.7% (meningiomas as proportion of all intracranial tumors (background))
  • count 15/19 (79%) (successful MEN-O establishment rate from meningioma samples)
  • count 11 grade 1, 4 grade 2 (WHO grade composition of successfully established MEN-O samples)
  • count 70-80% (typical proportion of meningiomas exhibiting some degree of PR positivity (background))
  • count N = 10-15 (number of MEN-Os collected per timepoint for representative sampling)
  • pvalue p < 0.05 (statistical significance threshold used for all analyses)

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.

This is a methods-development study establishing patient-derived meningioma organoids (MEN-Os) from 15/19 resected tumors, comparing organoids to their matched parent tissue using histopathological, immunohistochemical, and bulk RNA-sequencing (DESeq2, deconvolution, transcriptomic mutation calling) approaches. Group-level quantitative comparisons (e.g., marker positivity, proliferation) were analyzed with Student's t-test and Pearson's correlation in GraphPad Prism, with results reported as mean ± SEM and a significance threshold of p < 0.05. The study is explicitly described as a pilot protocol-validation study without a formal sample size calculation.

Replicationmixed Sample sizeExplicitly stated that no formal sample size calculation was performed, given the pilot nature of the study; 19 meningiomas collected, MEN-Os established from 15; 10-15 organoids sampled per timepoint; three technical replicates per RNA-seq library Groupsparent tumor tissue vs. corresponding patient-derived MEN-Os (also across culture timepoints 0/2/4 weeks) Pairingunclear Randomization/blindingnot stated for statistical comparisons (tumor grading was performed by an 'independent' neuropathologist, but blinding for quantitative/statistical analyses is not described) DispersionSEM
Statistical tests used
Test Applied to n Assumptions
Student's t-test comparisons between measurements (e.g., PR-positive cell counts, KI67 proliferation) between parent tumors and corresponding MEN-Os not specified per comparison; overall cohort N=15 organoid-generating samples not stated
Pearson's correlation sample correlation analysis (unspecified which variables) not stated not stated
DESeq2 differential gene expression analysis transcriptomic comparison between MEN-Os and parent tumor tissue not stated (bulk RNA-seq on parent tumors and matched MEN-Os after 4 weeks in culture) not stated
CIBERSORTx deconvolution inferring relative cell-type composition from bulk RNA-seq using a single-cell reference reference built from 6 previously sequenced meningiomas (GSE183655) not stated
Approaches that could also have been used
  • Student's t-test was applied across several distinct marker comparisons (e.g., PR positivity, KI67 proliferation) without a stated multiplicity-correction method.
    Could also: A one-way or two-way ANOVA with a post-hoc correction (e.g., Tukey HSD) or a Benjamini-Hochberg FDR adjustment across the family of comparisons — This would also help control the family-wise error rate or false discovery rate when several related comparisons are drawn from the same overall dataset.
  • Dispersion was reported as SEM.
    Could also: Standard deviation (SD) or a 95% confidence interval — SD directly conveys the spread of the underlying observations (SEM instead reflects precision of the mean), and a CI additionally communicates the plausible range for the true effect, which can be informative given the modest sample sizes (N=15-19) in this pilot study.
  • Comparisons between parent tumor tissue and its corresponding matched MEN-O appear to involve paired samples from the same patient.
    Could also: An explicitly paired t-test (or Wilcoxon signed-rank test for paired non-normal data) or a repeated-measures/mixed-effects model — Accounting for the paired/matched structure of parent-tumor-to-organoid data can increase statistical power by removing between-patient variability from the error term.
  • Parametric Student's t-test was used for comparisons such as marker-positive cell percentages, without stated assessment of normality.
    Could also: A non-parametric test such as the Mann-Whitney U test (unpaired) or Wilcoxon signed-rank test (paired) — These approaches do not assume a normal distribution and can be a useful complement when working with proportions or small-n biological samples.
  • The study explicitly notes that no formal sample size or power calculation was performed, consistent with its pilot design.
    Could also: A post hoc power calculation or reporting of standardized effect sizes (e.g., Cohen's d) — This would provide additional context on the precision and detectable effect magnitude of the comparisons, which can be useful for informing the design of future, larger-scale studies.
  • DESeq2 was used for differential expression analysis across many genes without explicit mention of the multiple-testing correction method in the text.
    Could also: Explicit reporting of Benjamini-Hochberg FDR-adjusted p-values (DESeq2's built-in default) alongside nominal p-values — Explicitly stating the adjusted-p threshold used to call genes differentially expressed makes clear how multiplicity across thousands of tested genes was addressed.
Software: GraphPad Prism 9.3 · R (DESeq2) 1.12 · R (PCAtools) 2.1 · Seurat V4 · SingleCellExperiment 1.28 · nf-core RNA-seq pipeline / FastQC / TrimGalore / STAR / Salmon FastQC 0.12, TrimGalore 0.6.10, STAR 2.7.11 · CIBERSORTx

What was reproduced

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

Scope — pmid-39941893

Paper: Zohdy YM et al. Patient-Derived Meningioma Organoids: A Reliable Model for Studying Human Tumor Pathophysiology. Cancers (Basel) 2025. PMID 39941893 · PMCID PMC11817449 · DOI 10.3390/cancers17030526.

Key metadata correction (vs harvested links)

The harvest filled data.json with GSE183655, but per the paper's Data Availability statement that is the reference single-cell dataset (six previously sequenced meningiomas) used only as the CIBERSORTx deconvolution reference. The study's OWN RNA-seq data is GSE287174 (BioProject PRJNA1211555, public, released 2025-02-11; 65 samples = 32 tumor + 33 organoid, with technical replicates). rnaseqmut (https://github.com/davidliwei/rnaseqmut) IS genuinely the tool the authors used for transcriptomic mutation calling — a third-party tool applied to the paper's own data (explicitly valid per BRIEF rule 2/P16).

Reported computational results

NOTE: the paper reports its bioinformatic results only as figures (heatmaps, scatter/PCA plots) — there are no numeric r-values, percentages, DEG counts or variance fractions printed in the text. This caps the achievable grade at "partial" (structural/quantitative reproduction with no exact printed number to match 1:1). Recorded honestly in claims.tsv.

Pipeline described in Methods

FastQC v0.12 → TrimGalore v0.6.10 → STAR v2.7.11 (align to hg38) → Salmon (quantification) → DESeq2 v1.12 (DE) + PCAtools v2.1 (PCA) in R. CIBERSORTx deconvolution (scRNA reference from GSE183655, Seurat v4 / SingleCellExperiment v1.28). rnaseqmut for SNP/indel calling from RNA-seq.

IN SCOPE (attempt — pipeline-derived, clearly specified, low compute)

id result figure pipeline data feasibility
C1 Transcriptomic similarity: organoids (MEN-Os) correlate strongly with their parent tumors (Pearson r) Fig 4b sample×sample Pearson on expression GSE287174_FPKM.csv.gz (shipped processed matrix) HIGH — small matrix, deterministic
C2 PCA: tumor vs organoid structure, low PC1/PC2 separation Fig 4a PCA (PCAtools) on FPKM GSE287174_FPKM.csv.gz HIGH — deterministic

STRETCH / last-20% (attempt only if cheap; may skip)

id result figure why hard
C3 SNP/indel overlap % between tumor & corresponding MEN-O Fig 5a needs raw FASTQ (GSE287174/SRA) → STAR → rnaseqmut for ≥1 pair; heavier compute, build rnaseqmut env

OUT OF SCOPE (not pipeline / not reproducible from public artifacts)

  • 79% (15/19) organoid culture success rate — wet-lab.
  • IHC / H&E / immunofluorescence / morphology, drug-response, proliferation — wet-lab.
  • CIBERSORTx deconvolution proportions (Fig 4d) — CIBERSORTx is a registration-gated web tool + requires rebuilding a scRNA signature matrix from GSE183655; out of 80/20 budget.

Strategy

Reproduce C1 + C2 from the paper's own shipped FPKM matrix on «our HPC»/«infra» (honest, low-compute, deterministic). Attempt C3 only if VPN/compute time allows. All data on «infra»; only small derived values returned to «host».

Figures / tables: Fig 4bFig 4aFig 5a
C1
Reported
Fig 4b: organoids (MEN-Os) correlate strongly with parent tumors; Pearson r heatmap, p<0.001 (figure only, no numeric r in text)
Reproduced
within-pair Pearson r mean=0.9421 (range 0.8987-0.9611, n=129; per-patient 0.92-0.96); cross-pair background r=0.918 (margin +0.024)
partial
C2
Reported
Fig 4a: PCA tumor vs organoid; low PC1/PC2 variability, right cultural shift on PC1, little-to-no PC2 variation (figure only, no variance % in text)
Reproduced
PC1=20.61%, PC2=14.07%; organoid PC1 centroid +36.9 vs tumor -35.8 (clear PC1 shift); PC2 +3.4 vs -3.5 (minimal) -- matches every stated qualitative point
partial
C3
Reported
Fig 5a: high SNP/indel overlap between tumors and corresponding MEN-Os via rnaseqmut (figure only, no % in text)
Reproduced
m.public.grade.not-run

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)
Scoring basis — itemised

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

Supporting (toward a concern)
Content-critical question only partially held
+2 pts
From: Q8 · Severity of the miss (overall human judgment) 🟡
Minor / cosmetic deviation
+1 pts
From: Q4 · Cause of the deviation 🟡
Minor / cosmetic deviation
+1 pts
From: Q2 · Endpoint comparability 🟡
Concordant (toward reproduced)
Code + data deposited & functional
-2 pts
From: Data & code availability Available & functional
Total score +2

Reproduction used the authors' own public processed matrix (GSE287174_FPKM) and confirmed the central transcriptomic-similarity conclusion: within-pair Pearson r̄=0.942 (0.92–0.96/patient) and a Fig 4a PCA showing the stated PC1 group shift with minimal PC2 separation. The only real limits are on the reporting/method side, not authorship integrity: the paper presents Fig 4a/4b/5a as figures with no printed numbers (so nothing is pinnable 1:1), PCA magnitudes are preprocessing-dependent, and C3 (Fig 5a rnaseqmut overlap) was not run. Severity is negligible and the core claim holds, but figure-only reporting plus an unrun claim make this a solid partial/yellow rather than a 1:1 green; the modest within-vs-cross margin (+0.024) is an honest nuance, not a contradiction.

🤝
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.

400.5 k
tokens (I/O) · 44.4 M incl. cache
77 min
runtime · 0.01 CPU-h
1.3 GB
peak RAM
1
HPC jobs
hummel
machine