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

Perturbation of placental protein glycosylation by endoplasmic reticulum stress promotes maladaptation of maternal hepatic glucose metabolism.

iScience · 2022
L1 71/100 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: Q5 · Derivability / plausibility 🟡
Content-critical question only partially held
+2 pts
From: Q7 · Core claim 🟡
Content-critical question only partially held
+2 pts
From: Q8 · Severity of the miss (overall human judgment) 🟡
Minor / cosmetic deviation
+1 pts
From: Q3 · Location of the main deviation 🟡
Minor / cosmetic deviation
+1 pts
From: Q4 · Cause of the deviation 🟡
Minor / cosmetic deviation
+1 pts
From: Q6 · Severity 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 +8
✓ What held up
  • Same input data as the authors
What did not (or only partly)
  • 🟡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
71/100
Reproducibility score
at the mean
vs. all fields · 1173 studies
🎯 Scores higher than 38% of all assessed papers rank 694 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

DESCRIBED WELL ENOUGH TO REPRODUCE; reproduced 1:1 on the paper's own deposited data with the paper's own methods, result = SAME numbers within a few percent (not byte-identical, by an explicitly documented method deviation). Yung et al., iScience 2022: bulk RNA-seq of BeWo trophoblast cells +/- thapsigargin (ER stress), 5 paired Control/Tg replicates. The analysis-ready count matrix (deseq2.dds.RData, 60504 genes) is NOT deposited in ArrayExpress E-MTAB-10943 (only raw reads + IDF/SDRF) and NOT shipped in the GitHub repo, so counts were regenerated from the 31 raw single-end FASTQs (ENA ERP131881, ~20 GB, all gzip-validated; 31 lanes mapped to 10 samples via SLX filenames, matching the repo SampleTable). Pipeline: Salmon 1.5.2 (decoy-aware, mapping-based selective alignment) vs Ensembl GRCh38.104 cdna+ncrna -> tximport -> DESeq2 ~Pairs+condition + apeglm. RESULTS: C2 genes-passing-filter 23395 vs 23426 (within-tol, -0.13%); C3 non-NA-padj 20978 vs paper's two inconsistent figures 21076(code)/20701(prose) (within-tol, lands between them); C4 significant DEGs 3091 (1673 up/1418 down) vs 3194 (1712/1482) (within-tol, -3.2%, same up/down split); C5 GO:0006486 protein-glycosylation 103 of 153 changed vs paper 93 of 173 (partial -- same order of magnitude, annotation-source drift); C1 total genes 57424 vs 60504 (partial, -5.1%, mapping-based cdna+ncrna index carries fewer genes than nf-core's STAR full-GTF transcriptome). OVERALL: the headline differential-expression pipeline reproduces faithfully (C2-C4 within-tol); C1/C5 are partial for fully-explained method/annotation reasons. AUDITABLE FINDINGS: (a) the paper's Methods prose (20701 non-NA padj; 23426 filtered) contradicts its own shipped R script (code comment 21076; no normalised-count pre-filter before DESeq) -- an internal prose-vs-code inconsistency, flagged for the reviewer, NOT a fabrication; (b) no evidence of fabrication -- every graded number regenerates within a few percent from the deposited reads. METHOD/ENV DEVIATIONS (honest): Salmon mapping-based vs nf-core star_salmon alignment-based; R 4.5.3/DESeq2 1.50.2 vs paper R 4.0.2/DESeq2 1.30.1; org.Hs.eg.db current vs Ensembl-104 for GO. NOT ATTEMPTED (out of scope): full STAR/nf-core orchestration (Singularity unavailable); all wet-lab results (glycomics, TMT-LC/MS proteomics, mouse Sp-Perk-/- model, hCG/PlGF/VEGF bioactivity, lectin work) which are non-computational.

💻 Code ↗ 🗄 Data: E-MTAB-10943

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 Score 50
    assessed: 2026-06-15 ⛓ 7a491fc8b84a
✎ 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
no human curator yet
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 authors hypothesized that placental ER stress, as seen in early-onset pre-eclampsia, compromises protein glycosylation and thereby reduces the bioactivity of placental hormones, causing maladaptive maternal physiological/metabolic changes that predispose the mother to later metabolic disease.

Core claims
  • ER stress reduces the complexity and sialylation of trophoblast protein N-glycosylation finding
  • ER stress-induced aberrant glycosylation of VEGFA reduces its bioactivity (failure to activate VEGFR2) finding
  • ER stress alters expression of 66 of 146 genes annotated with 'protein glycosylation' and reduces sialyltransferase expression finding
  • Mouse junctional zone placental explants under ER stress secrete mis-glycosylated glycoproteins (e.g., CEACAM11/12) finding
  • Pregnant mice with junctional zone-specific placental ER stress show reduced blood glucose, anomalous hepatic glucose metabolism, increased cellular stress, and elevated DNMT3A finding
  • Women with early-onset pre-eclampsia show aberrant glycosylation of placental pregnancy-specific glycoproteins finding
  • Development of a novel ex vivo mouse junctional zone placental explant culture model method
  • Eif2s1 tm1RjK mutant MEFs serve as a genetic model of chronic ER stress independent of pharmacological confounds resource
Experimental setups
Assay System Perturbation Readout Platform
MALDI-TOF MS glycomics BeWo trophoblast-like cells (conditioned media) thapsigargin (Tg) N-glycan subtype distribution and sialylation MALDI-TOF mass spectrometry
Western blot BeWo cells (cell lysate and conditioned media) thapsigargin (Tg) or tunicamycin (Tm) hCGβ glycosylation pattern
Western blot BeWo cells (cell lysate and conditioned media) thapsigargin (Tg) or tunicamycin (Tm) VEGFA glycosylation/isoform pattern
RT-PCR and Western blot Eif2s1 tm1RjK mutant mouse embryonic fibroblasts (MEFs) genetic ER stress (Eif2s1 S51A knock-in) Vegfa transcript and secreted VEGFA protein pattern
Receptor phosphorylation bioassay human umbilical vein endothelial cells (HUVECs) conditioned media from wt or Eif2s1 mutant MEFs VEGFR2 (KDR) phosphorylation
bulk RNA-seq BeWo cells thapsigargin (Tg) differential gene expression, GO/GSEA enrichment DESeq2, gprofiler2, GSEA
RT-qPCR mouse placenta (junctional zone and labyrinthine zone) none (tissue characterization) Tpbpa/Gcm1 marker gene expression for zone purity
Ex vivo explant culture with Western blot mouse placental junctional zone (Jz) explants thapsigargin (Tg) ER stress markers ATF4, GRP78, XBP1
Key results
  • Glycan subtype proportions (oligomannose/hybrid/complex) shifted from 26%/5%/69% (control) to 51%/19%/30% (Tg-treated) BeWo secretome 26/5/69% to 51/19/30%
  • Tg reduced relative abundance of fully sialylated N-glycan species vs partially/un-sialylated species
  • Secreted hCGβ glycosylation markedly decreased after Tg and Tm treatment
  • Conditioned media from Eif2s1 mutant MEFs failed to induce VEGFR2 phosphorylation in HUVECs, unlike wild-type media or recombinant VEGFA
  • 1712 transcripts increased and 1482 decreased in BeWo cells after Tg treatment padj<0.05, FC≥2
  • 66 of 146 'protein glycosylation' GO-annotated genes were differentially expressed after Tg, a significant over-representation Fisher exact p=2.2×10^-16
  • Sialyltransferase mRNAs ST3GAL4, ST3GAL6, ST6GAL1, ST8SIA4 were reduced in Tg-treated BeWo cells
  • Tg treatment induced a 2.6-fold increase in ATF4 in Jz explants after 48h, indicating mild ER stress 2.6-fold, p=0.005
Key statistics
  • pvalue P adj = 8.4 × 10^-11 (GO:0044786 cell cycle DNA replication enrichment in down-regulated genes)
  • pvalue P adj = 5.2 × 10^-20 (GO:0034976 response to ER stress enrichment in up-regulated genes)
  • count 1712 up, 1482 down transcripts (DESeq2 DEGs after Tg treatment, padj<0.05, |FC|≥2)
  • count 66 of 146 genes (protein glycosylation GO:0006486 genes differentially expressed after Tg; Fisher exact p=2.2×10^-16)
  • fold_change 2.6-fold increase, p=0.005 (ATF4 induction in Jz explants after Tg treatment)
  • mean Jz contamination 11.9% ± 4.4%; Lz contamination 1.2% ± 0.4% (purity assessment of dissected Jz and Lz placental regions)
  • pvalue CEACAM11 spot 1503 +13% (p=0.06), spot 1553 +11.8% (p=0.02), spot 1549 -16.5% (p=0.046), spot 1546 -20.4% (p=0.027) (DIGE spot intensity changes after Tg in Jz explant conditioned media)
  • other GSEA GOBP_ER_Unfolded Protein Response NES = -2.68, p=0.000, FDR=0.000 (top GSEA enrichment result after Tg treatment)

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 combined in vitro (BeWo trophoblast cells), ex vivo (mouse junctional zone explants), in vivo (transgenic Jz-Perk−/− mice), and human clinical samples to test whether placental ER stress perturbs protein glycosylation and maternal glucose metabolism. Transcriptomic changes were assessed by RNA-Seq with DESeq2 differential expression (Padj<0.05, |FC|≥2) and GSEA/g:Profiler pathway enrichment; individual sialyltransferase comparisons used paired t-tests. Glycomic profiles were described proportionally from MALDI-TOF MS data, 2D-DIGE assessed protein spot shifts with p-value thresholds, and Western blotting was used qualitatively. Results were reported as fold-changes, adjusted p-values, normalized enrichment scores, and mean ± SD.

Replicationmixed Sample sizeRNA-Seq: 5 independent replicate pairs (BeWo cells); glycomic analysis: 2 independent experiments; Jz purity RT-qPCR: n=4; sample sizes for mouse in vivo and human ePE cohorts not reported in available text GroupsTg- or Tm-treated vs. untreated BeWo cells; Eif2s1tm1RjK mutant vs. wild-type MEFs; Jz explants ±Tg; transgenic Jz-Perk−/− mice vs. controls; women with ePE vs. uncomplicated pregnancies Pairingmixed Randomization/blindingnot stated DispersionSD Exact p-valuesyes Effect sizesyes Confidence intervalsno Multiplicity correctionBenjamini-Hochberg FDR (DESeq2 Padj by default; GSEA FDR q-values; gprofiler2 Padj); no correction stated for the four individual sialyltransferase paired t-tests or for multiple 2D-DIGE spot comparisons
Statistical tests used
Test Applied to n Assumptions
DESeq2 Wald test RNA-Seq differential expression, Tg-treated vs. untreated BeWo cells (Figure 2, Tables S2, S7) 5 independent replicate pairs not stated
Fisher's exact test (two-sided) Over-representation of differentially expressed genes among 146 GO:0006486 'protein glycosylation' annotated genes 146 annotated genes present in dataset not stated
Gene Set Enrichment Analysis (GSEA) with FDR q-value GO biological process pathway enrichment in RNA-Seq data, including ER stress, mannose trimming, and protein deglycosylation gene sets (Figure 2B, Figure S3) 5 independent replicate pairs not stated
g:Profiler (gprofiler2) over-representation analysis GO biological process term enrichment of up- and down-regulated DEGs after Tg treatment (Figure 2A) 5 independent replicate pairs not stated
Paired t-test Individual sialyltransferase mRNA normalized counts (ST3GAL4, ST3GAL6, ST6GAL1, ST8SIA4) in Tg vs. control BeWo cells (Figure 2D) n = 5 pairs not stated
Statistical test for 2D-DIGE spot intensity ratios (specific test not named) CEACAM11 and CEACAM12 protein spot intensity changes in Jz explant conditioned media (Figure 3C) not stated
Approaches that could also have been used
  • Glycomic proportions (oligomannose, hybrid, complex glycans) were compared between conditions using pie charts derived from 2 independent experiments, without a formal statistical test or uncertainty estimate
    Could also: A chi-squared test, Dirichlet-multinomial model, or compositional data analysis (e.g., ANOVA on isometric log-ratio transformed proportions) across a larger number of biological replicates could also be applied to glycan proportion data — Formal compositional testing would yield p-values and interval estimates for the observed proportional shifts, making it possible to distinguish signal from sampling variability and to report uncertainty alongside the point estimates
  • Four sialyltransferase genes were each tested individually with separate paired t-tests at p<0.05 (Figure 2D)
    Could also: A linear mixed model treating gene identity as a within-subject factor, or Bonferroni/Benjamini-Hochberg correction applied across the four tests, could also be used — When several related hypotheses are tested simultaneously, multiplicity adjustment is a standard approach to control the family-wise error rate or false discovery rate, consistent with the correction applied to the broader RNA-Seq analysis
  • 2D-DIGE spot intensity changes for CEACAM11 and CEACAM12 were reported at unadjusted p<0.05 across multiple spots (Figure 3C)
    Could also: FDR correction (e.g., Benjamini-Hochberg) across all 19 selected protein spots would also be applicable — Applying FDR adjustment when testing multiple protein spots in parallel limits the expected proportion of false discoveries and is consistent with the approach used in the RNA-Seq component of the same study
  • VEGFA bioactivity was evaluated qualitatively by Western blot (presence/absence of phospho-VEGFR2 bands) without stated replication number or quantification (Figure 1E)
    Could also: Densitometric quantification of band intensities across independently replicated experiments, followed by a t-test or Wilcoxon test on the ratios, could also provide a quantitative effect estimate — Quantitative densitometry with biological replication enables estimation of effect magnitude and variability, supporting more precise inference about the functional consequence of mis-glycosylation
  • Results for continuous outcomes were summarized with mean ± SD (e.g., Figure 2D, n=5)
    Could also: 95% confidence intervals could also be reported alongside or instead of SD, particularly for small n — Confidence intervals convey both the estimated magnitude and its precision in a single quantity, and are recommended by many reporting guidelines (e.g., ARRIVE, Nature Methods) as they directly address the uncertainty in the population-level estimate
  • DEG identification applied a combined threshold of Padj<0.05 and a hard fold-change cutoff (≥2 for global DEGs; ≥1.5 for glycosylation genes)
    Could also: Shrinkage-based effect-size estimation (e.g., DESeq2 lfcShrink with apeglm or ashr) without a hard FC threshold, presented on a volcano or MA plot, could also be used — Hard fold-change thresholds can be dominated by low-count gene noise; shrunken log-fold-change estimates reduce noise-driven extreme values for low-count genes and are increasingly recommended for transparent RNA-Seq reporting
Software: DESeq2 · gprofiler2 · GSEA · MALDI-TOF MS (instrument/analysis software unspecified) · LC-MS/MS (instrument/analysis software unspecified)

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
29
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.

RRID:AB_2315112 RRID in Article (http://semanticscience.org/resource/SIO_001029)
also used by 2 papers:
RRID:AB_2250373 RRID in Article (http://semanticscience.org/resource/SIO_001029)
also used by 1 paper:
RRID:AB_331659 RRID in Article (http://semanticscience.org/resource/SIO_001029)
also used by 1 paper:
E-MTAB-10943 ArrayExpress in Methods (http://purl.org/orb/Methods)
no other assessed paper uses this yet
GO:0006486 Gene Ontology (GO) in Results (http://purl.org/orb/Results)
no other assessed paper uses this yet
Q00731 UniProt in Results (http://purl.org/orb/Results)
no other assessed paper uses this yet
RRID:AB_2212507 RRID in Article (http://semanticscience.org/resource/SIO_001029)
no other assessed paper uses this yet
RRID:AB_2212642 RRID in Article (http://semanticscience.org/resource/SIO_001029)
no other assessed paper uses this yet
RRID:AB_329830 RRID in Article (http://semanticscience.org/resource/SIO_001029)
no other assessed paper uses this yet
RRID:AB_330330 RRID in Article (http://semanticscience.org/resource/SIO_001029)
no other assessed paper uses this yet
RRID:AB_330331 RRID in Article (http://semanticscience.org/resource/SIO_001029)
no other assessed paper uses this yet
RRID:AB_330713 RRID in Article (http://semanticscience.org/resource/SIO_001029)
no other assessed paper uses this yet
RRID:AB_330745 RRID in Article (http://semanticscience.org/resource/SIO_001029)
no other assessed paper uses this yet
RRID:AB_331367 RRID in Article (http://semanticscience.org/resource/SIO_001029)
no other assessed paper uses this yet
RRID:AB_331641 RRID in Article (http://semanticscience.org/resource/SIO_001029)
no other assessed paper uses this yet
RRID:AB_476697 RRID in Article (http://semanticscience.org/resource/SIO_001029)
no other assessed paper uses this yet
RRID:AB_490890 RRID in Article (http://semanticscience.org/resource/SIO_001029)
no other assessed paper uses this yet
RRID:IMSR_JAX:017601 RRID in Article (http://semanticscience.org/resource/SIO_001029)
no other assessed paper uses this yet
RRID:IMSR_JAX:023066 RRID in Article (http://semanticscience.org/resource/SIO_001029)
no other assessed paper uses this yet
RRID:MGI:2159769 RRID in Article (http://semanticscience.org/resource/SIO_001029)
no other assessed paper uses this yet

What was reproduced

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

Scope — pmid-36660474

Paper: Yung et al., "Perturbation of placental protein glycosylation by ER stress promotes maladaptation of maternal hepatic glucose metabolism." iScience 2022. PMID 36660474 · PMC9843443 · DOI 10.1016/j.isci.2022.105911. Code: https://github.com/CTR-BFX/Yung_Charnock-Jones (R analysis script + shipped figures/tables; authors' own code). Data: ArrayExpress E-MTAB-10943 = ENA study ERP131881 (31 single-end FASTQs, ~21 GB, 10 biological samples × multiple sequencing lanes).

The paper's pipeline (as described in Methods + README + R script)

  1. Raw FASTQ → gene counts: nf-core/rnaseq v3.2 (--aligner star_salmon), Nextflow 21.05.0.edge, Ensembl GRCh38 release 104. Produces a salmon gene-level count object deseq2.dds.RData (60504 genes).
  2. Differential expression: R 4.0.2, DESeq2 1.30.1, design ~Pairs+condition (5 paired control/Tg reps), apeglm LFC shrinkage. Filtering, DEG calling.
  3. Downstream: GO/GSEA (gprofiler2, clusterProfiler), heatmaps (ComplexHeatmap), PCA — all consume the DESeq2 output.

In scope (pipeline-derived, attempted)

The count matrix deseq2.dds.RData is not shipped in the repo and no processed matrix is attached to E-MTAB-10943 (only IDF/SDRF metadata). So the counts must be regenerated from the raw FASTQs. We reproduce the quantification + differential-expression core, which underlies every headline number, by running the paper's own quantifier (Salmon, the salmon half of star_salmon, pinned to v1.5.2 as in nf-core/rnaseq 3.2) directly on the ENA FASTQs against Ensembl GRCh38.104 (decoy-aware index), then tximport → DESeq2 with the identical design and thresholds.

Reproduced data points (compared in claims.tsv):

  • C1 total genes quantified (paper 60504)
  • C2 genes passing low-count filter, normalised rowSums>10 (paper methods: 23426)
  • C3 genes with non-NA padj after DESeq2 (paper code comment: 21076; prose: 20701)
  • C4 significant DEGs |log2FC|≥1 & padj<0.05 (paper 3194 = 1712 up + 1482 down)
  • C5 GO:0006486 "protein glycosylation": genes changed (padj<0.05) of total annotated (paper abstract: 93 of 173)

Deviation from the original pipeline (honest 1:1 caveat)

nf-core star_salmon runs STAR genome alignment then Salmon in alignment-based mode on the transcriptome BAM; we run Salmon in mapping-based (selective- alignment, decoy-aware) mode directly on the reads. Same tool, same data, same reference annotation, same DESeq2 design — but counts are expected to be very close rather than byte-identical, so integer-level agreement on C2–C5 is graded within-tol/partial, not exact. GO annotations (C5) are queried from current Ensembl biomaRt and will drift slightly from release-104-era annotations.

Out of scope (not attempted, why)

  • Full STAR genome alignment + nf-core/rnaseq orchestration: the heavy ~20%; Singularity is unavailable on «our HPC» and a 20+-process conda-per-process Nextflow run is high-risk for marginal fidelity gain over Salmon counts.
  • Wet-lab results (glycomics, TMT-LC/MS proteomics, mouse Sp-Perk−/− model, hCG/PlGF/VEGF bioactivity, lectin isolation): non-computational, out of scope.
  • Downstream GSEA/GO-enrichment plots, heatmaps, WikiPathways/MSigDb/cellMarker enrichments: derivative of the DEG list; not separately graded (80/20).
C1
Reported
60504 genes quantified (star_salmon, Ensembl GRCh38.104)
Reproduced
57424
partial
C2
Reported
23426 genes pass low-count filter (normalised counts sum>10)
Reproduced
23395
within tolerance
C3
Reported
20701 (prose) / 21076 (shipped-code comment) genes with non-NA padj
Reproduced
20978
within tolerance
C4
Reported
3194 significant DEGs (1712 up + 1482 down), |log2FC|>=1 & padj<0.05
Reproduced
3091 (1673 up + 1418 down)
within tolerance
C5
Reported
GO:0006486 protein glycosylation: 93 of 173 genes changed (padj<0.05)
Reproduced
103 of 153
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 71/100

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: Q5 · Derivability / plausibility 🟡
Content-critical question only partially held
+2 pts
From: Q7 · Core claim 🟡
Content-critical question only partially held
+2 pts
From: Q8 · Severity of the miss (overall human judgment) 🟡
Minor / cosmetic deviation
+1 pts
From: Q3 · Location of the main deviation 🟡
Minor / cosmetic deviation
+1 pts
From: Q4 · Cause of the deviation 🟡
Minor / cosmetic deviation
+1 pts
From: Q6 · Severity 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 +8

This is a careful 1:1 reproduction attempt on the authors' own data — raw FASTQs (ENA ERP131881) are public and were mapped exactly to the repo SampleTable, but the processed count matrix is not deposited, so counts had to be regenerated. The run finalized PARTIAL with no reproduced integers (SLURM job still index-building at cutoff), so C1-C5 cannot yet be graded and the central (largely wet-lab) conclusion can only be marked limited, not confirmed or refuted. The one concrete established finding is an authors'-side prose-vs-code inconsistency (Methods 23426 input / 20701 retained vs shipped-code 21076 and no normalised-count pre-filter), affecting C2/C3. Net: solid, honest setup with explainable deviations (data-availability + our Salmon-vs-STAR method choice) and no fabrication signal — overall yellow.

🤝
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.

354.1 k
tokens (I/O) · 31.2 M incl. cache
207 min
runtime · 9.44 CPU-h
20.4 GB
peak RAM
4 (2 failed)
HPC jobs
hummel
machine