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

Meta-analysis of gene expression profiles of lean and obese PCOS to identify differentially regulated pathways and risk of comorbidities.

Comput Struct Biotechnol J · 2020
L1 78/100 3/4
Why this verdict

The main results reproduced: recomputed values matched the published ones within tolerance.

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: Q3 · Location of the main deviation 🟡
Minor / cosmetic deviation
+1 pts
From: Q4 · Cause of the deviation 🟡
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
  • Reported values were directly comparable
  • Reported values are derivable from the shared data
  • Any deviation was negligible
  • The central claim held under reproduction
What did not (or only partly)
  • 🟡A deviation arose in the data or preprocessing
  • 🟡A deviation was attributed to the published material
  • 🟡Overall, the reproduction showed a material discrepancy
How its reproducibility compares
78/100
Reproducibility score
at the mean
vs. all fields · 1173 studies
🎯 Scores higher than 51% of all assessed papers rank 533 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 the DIFFERENTIAL-EXPRESSION stage 1:1. Ran the authors' own AffyOldLimma.R (faithful port: affy::justRMA + BioMart symbol map + collapse-dup-by-maxSD + limma ModT) on the paper's public GEO CELs for the two unambiguous LEAN GPL570 datasets — GSE98421 (the brief's named accession; subcutaneous adipose, 4v4) and the GSE10946 lean subset (cumulus, 6 nonPCOS-Lean vs 5 PCOS-Lean, exactly matching the paper's 6+5). RESULT: the paper's crisp gene-level claim reproduces EXACTLY — all 7 named commonly-dysregulated lean-PCOS genes (PRRT1, SLITRK4, CRHBP, HAPLN1, SRGN, EREG, WNT5A) are significant DEGs in BOTH lean tissues (intersection of 9, containing precisely those 7). The paper's asymmetric-threshold down-skew (4891 down vs 123 up) also reproduces per tissue (150/2, 892/16). Auditability flag (not fabrication): the reported DEGs are nominal raw-p<0.05 and none survive BH correction in these n=4-6/arm studies. NOT ATTEMPTED (the under-specified / GUI-bound / unshipped-data ~20%): (a) the full 8-dataset aggregate (5014/4224 DEGs, tissue split) — needs GPL6244-ST with ambiguous 16-sample selection (GSE98595), a non-Affy GPL15362 custom array (GSE43264), and GPL96 where the shipped script's BioMart filter is hardcoded to U133-Plus-2; (b) pathway gene-set counts 86/1031/159 and the 4 themes — GSEA v3.0 preranked + interactive Cytoscape 3.6.1 Enrichment Map GUI with the Bader-lab Nov-2018 GMT; (c) the 7 transcription factors (downstream of b); (d) GSVA disease/ICD-11 mapping (222/192/127, 37 visual, 136 PCOS genes) — GSVA-HeatMap.R reads gene-set files not shipped in the repo (hardcoded G:/ paths). Verdict: PARTIAL but high-confidence 1:1 on the deterministic shipped-code DE stage incl. the paper's headline gene signature; no fabrication detected.

💻 Code ↗ 🗄 Data: GSE98421

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 78
    assessed: 2026-06-15 ⛓ 39022467b0aa
✎ 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
2026-06-15
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: opus
Founding hypothesis

Lean and obese PCOS may have divergent pathophysiology; this meta-analysis systematically compares gene expression profiles of lean versus obese PCOS to identify differentially expressed genes, enriched pathways, and associated comorbidities across tissue types.

Core claims
  • The majority of differentially expressed genes in PCOS are downregulated regardless of tissue type and phenotype finding
  • Ovarian and endometrial tissues share many commonly dysregulated genes, suggesting shared PCOS pathophysiology mechanisms across tissues finding
  • Pathways for cell-motility/immune response, FAK, ERBB1/PDGFRB signaling are downregulated in lean PCOS but upregulated in obese PCOS, while mitochondrial gene expression is upregulated in lean and downregulated in obese PCOS mechanism
  • The gene-disease network is denser with a higher comorbidity score (GDS) for obese PCOS compared to lean PCOS finding
  • DEGs map to developmental, metabolic, nervous, and visual system diseases, suggesting comorbidity hypotheses for PCOS finding
  • A meta-analysis approach normalizing and analyzing each case-control GEO study individually then comparing DEG/pathway results across studies method
  • Seven genes (PRRT1, SLITRK4, CRHBP, HAPLN1, SRGN, EREG, WNT5A) are commonly dysregulated in lean PCOS across cumulus cells and subcutaneous adipose tissue (GPL570) finding
  • WNT5A is upregulated in cumulus granulosa cells of lean PCOS, consistent with prior qPCR reports finding
Experimental setups
Assay System Perturbation Readout Platform
Microarray gene expression (meta-analysis) Lutein granulosa cells (GSE98595, lean PCOS) none (PCOS cases vs controls) differentially expressed genes (logFC, p-value) Affymetrix Human Gene 1.0 ST Array (GPL6244)
Microarray gene expression (meta-analysis) Subcutaneous adipose tissue (GSE98421 lean; GSE43264 obese) none (PCOS cases vs controls) differentially expressed genes Affymetrix Human Genome U133 Plus 2.0 Array (GPL570); NuGO array NuGO_Hs1a520180 (GPL15362)
Microarray gene expression (meta-analysis) Cumulus cells (GSE10946, lean and obese PCOS) none (PCOS cases vs controls) differentially expressed genes Affymetrix Human Genome U133 Plus 2.0 Array (GPL570)
Microarray gene expression (meta-analysis) Skeletal muscle (GSE6798, obese PCOS) none (PCOS cases vs controls) differentially expressed genes Affymetrix Human Genome U133 Plus 2.0 Array (GPL570)
Microarray gene expression (meta-analysis) Metaphase II oocyte (GSE5850, obese PCOS) none (PCOS cases vs controls) differentially expressed genes Affymetrix Human Genome U133 Plus 2.0 Array (GPL570)
Microarray gene expression (meta-analysis) Endometrial epithelial/endothelial/stromal fibroblast/mesenchymal stem cells (GSE48301, obese PCOS) none (PCOS cases vs controls) differentially expressed genes Affymetrix Human Gene 1.0 ST Array (GPL6244)
Microarray gene expression (meta-analysis) Omental adipose tissue (GSE5090, obese PCOS) none (PCOS cases vs controls) differentially expressed genes Affymetrix Human Genome U133A Array (GPL96)
Pathway enrichment analysis (GSEA) and gene set variation analysis (GSVA) GPL570 ovarian/endometrial/adipose/skeletal tissues; KEGG disease gene sets none (in silico) normalized enrichment scores, enriched pathways, gene-disease association score (GDS) GSEA v3.0 (Broad), Cytoscape v3.6.1, GSVA/complexheatmap R packages, KEGG disease DB Release 88.2
Key results
  • 5014 statistically significant DEGs (unique = 4224) identified across eight GEO datasets 5014 (4224 unique)
  • 4891 DEGs downregulated (unique=4101) vs 123 upregulated (unique=96), majority downregulated 4891 down vs 123 up
  • Ovarian and endometrial tissues shared 181 (4.3%) commonly dysregulated genes, the maximum shared 181 genes (4.3%)
  • Six genes (GPX7, SERPINI1, TMEM256, SVIP, MAT2A, SRGN) commonly dysregulated in ovarian, endometrial and adipose tissues 6 genes
  • Pathway gene sets: 86 (6.7%) common, 1031 (80.8%) unique to lean, 159 (12.5%) unique to obese PCOS 86 common / 1031 lean / 159 obese
  • DEGs from obese PCOS and endometrial tissues displayed maximum gene-disease association score (GDS)
  • DEGs mapped to developmental anomalies (222 genes), metabolic disorders (192), nervous system (127), visual system (37) 222/192/127/37 genes
  • No commonly dysregulated genes across all platforms/tissues; 7 genes common in lean PCOS limited to cumulus cells and subcutaneous adipose (GPL570) 7 genes
Key statistics
  • count 5014 DEGs (4224 unique) (total statistically significant DEGs across eight GEO datasets)
  • count 4891 downregulated (4101 unique), 123 upregulated (96 unique) (DEG direction breakdown)
  • pvalue p < 0.05; logFC > 2 (up) and < -0.5 (down) (DEG significance thresholds in limma)
  • count 1284 ovarian-exclusive, 2473 endometrial-exclusive, 202 adipose-exclusive, 7 skeletal-exclusive DEGs (tissue-exclusive DEG distribution)
  • count 86 (6.7%) common, 1031 (80.8%) lean-unique, 159 (12.5%) obese-unique pathway gene sets (GPL570 pathway enrichment comparison)
  • count 222 developmental, 192 metabolic, 127 nervous, 37 visual system genes (DEGs mapped to ICD-11 disease categories)
  • other 2000 permutations; gene set Max=200, Min=10; weighted statistic p=1; FDR q=0.1 (GSEA/Enrichment Map parameters)
  • count 136 of 4224 unique DEGs have established PCOS association in literature (literature validation of DEGs)

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 paper used a study-level meta-analysis approach in which eight GEO microarray datasets of lean and obese PCOS were each independently preprocessed (RMA normalization) and analyzed for differentially expressed genes (DEGs) via limma moderated t-statistics, then DEG lists were compared across studies by intersection rather than by pooled effect estimation. Pathway enrichment was assessed per dataset with GSEA (permutation-based FDR), and results were visualized and clustered with Cytoscape EnrichmentMap. Comorbidity burden was quantified with a custom gene-disease association score (GDS) mapping DEGs to KEGG disease categories, and GSVA was used to score gene-set variation across tissue types.

Replicationbiological Sample sizePer-dataset sample sizes given in Table 1 (3–16 per group); no formal a priori power calculation described GroupsPCOS cases vs. non-PCOS controls within each dataset; lean PCOS datasets vs. obese PCOS datasets across datasets Pairingunpaired Randomization/blindingnot stated DispersionSD Exact p-valuesno Effect sizesyes Confidence intervalsno Multiplicity correctionGSEA permutation-based FDR (q-value cutoff 0.1) for pathway enrichment; no multiple-testing correction stated for per-gene limma tests
Statistical tests used
Test Applied to n Assumptions
limma moderated t-statistic (empirical Bayes shrinkage of variance estimates) Identification of DEGs in each of the eight GEO datasets individually (PCOS cases vs. controls) Varies by dataset: 3–16 controls and 3–16 PCOS subjects per dataset (see Table 1) not stated
GSEA weighted enrichment statistic (permutation-based, 2000 permutations) Pathway gene-set enrichment analysis for each dataset; FDR q-value < 0.1 and p-value < 1.0 cutoffs applied Gene sets sized 10–200 genes; based on same per-dataset sample sizes as DEG analysis not stated
Gene Set Variation Analysis (GSVA) Assessment of ICD-11 disease-associated gene-set variation across ovarian tissue types (GPL570 platform only) for lean and obese PCOS GPL570 ovarian tissue datasets only; exact combined n not restated at this stage not stated
Custom gene-disease association score (GDS): (DEGs mapped to disease / total genes mapped to disease) × 100 Comorbidity analysis mapping DEGs to KEGG disease categories by ICD-11 classification 4224 unique DEGs mapped against KEGG disease database Release 88.2 na
Approaches that could also have been used
  • DEGs were filtered using an unadjusted p-value threshold (p < 0.05) combined with logFC cutoffs across tens of thousands of probes per array
    Could also: Apply Benjamini-Hochberg FDR adjustment (adjusted p < 0.05 or 0.10) to the per-gene limma output before applying logFC filters — With 17,000–54,000 probes tested per array, BH-FDR correction is a widely adopted approach to limit the expected proportion of false discoveries among reported DEGs; reporting both adjusted and unadjusted results would allow readers to evaluate the sensitivity-specificity trade-off, especially important here given small per-dataset sample sizes
  • Cross-dataset meta-analysis was performed by comparing lists of significant DEGs across studies (intersection / vote-counting approach)
    Could also: Use a formal statistical meta-analysis of effect sizes, such as Fisher's combined p-value method, a random-effects model (e.g., metaMA or RankProd), or a combined z-score approach — Pooled effect-size meta-analysis quantifies the direction and magnitude of expression change consistently across studies, propagates per-study uncertainty into a single estimate, and can formally test for heterogeneity between datasets — information that intersection of binary DEG lists does not provide
  • Asymmetric logFC thresholds were applied: logFC > 2 for upregulation and logFC < −0.5 for downregulation
    Could also: Apply a symmetric absolute logFC threshold (e.g., |logFC| ≥ 1) or rank-based filtering that does not require a directionally asymmetric cutoff — Symmetric or |logFC|-based thresholds treat up- and downregulation equivalently, which simplifies interpretation and is the more common convention in the microarray literature; asymmetric thresholds warrant explicit biological justification so readers understand why downregulated genes were captured at a lower magnitude
  • Pathway enrichment used the ranked-list GSEA method applied to each individual dataset
    Could also: Use Over-Representation Analysis (ORA) via Fisher's exact test or a hypergeometric test on the filtered DEG lists, or apply a cross-dataset enrichment aggregation (e.g., combining per-dataset enrichment scores before ranking pathways) — ORA on the intersection DEG lists directly connects the DEG-comparison step to enrichment, and is more interpretable when n is small; cross-dataset aggregation of enrichment scores would make the pathway-level meta-analysis as explicit as the gene-level one
  • Comorbidity burden was quantified with a custom GDS formula (proportion of disease-associated genes that appear in the DEG list)
    Could also: Apply a hypergeometric test or Fisher's exact test to assess whether overlap between DEGs and disease gene sets exceeds chance, using established gene-disease databases such as DisGeNET or the DISEASES resource — A statistical test provides a p-value and confidence interval for each gene-disease association, allowing formal comparison of comorbidity enrichment between lean and obese PCOS rather than a descriptive proportion score; established databases also provide evidence-weighted gene-disease links
  • Datasets from different microarray platforms and diverse tissue types were kept fully separate throughout the analysis
    Could also: Apply cross-platform batch correction (e.g., ComBat from the sva R package) to harmonize datasets sharing the same platform after normalization, then model tissue type as a covariate — Batch correction can reduce platform-driven variance and enable direct cross-dataset comparisons; it is optional but increasingly common when merging microarray studies, and would allow a single integrated DEG or enrichment analysis rather than purely list-level comparisons
Software: R/limma · R/affy · R/oligo · R/biomaRt · R/GSVA · R/ComplexHeatmap · GSEA (Broad Institute Java desktop application) 3.0 · Cytoscape with EnrichmentMap and AutoAnnotate plugins 3.6.1 · KEGG disease database Release 88.2

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
26
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.

GPL15362 GEO in Methods (http://purl.org/orb/Methods)
no other assessed paper uses this yet
GSE10946 GEO in Results (http://purl.org/orb/Results)
no other assessed paper uses this yet
GSE43264 GEO in Methods (http://purl.org/orb/Methods)
no other assessed paper uses this yet
GSE48301 GEO in Methods (http://purl.org/orb/Methods)
no other assessed paper uses this yet
GSE5090 GEO in Methods (http://purl.org/orb/Methods)
no other assessed paper uses this yet
GSE5850 GEO in Methods (http://purl.org/orb/Methods)
no other assessed paper uses this yet
GSE6798 GEO in Results (http://purl.org/orb/Results)
no other assessed paper uses this yet
GSE98421 GEO in Results (http://purl.org/orb/Results)
no other assessed paper uses this yet
GSE98595 GEO in Methods (http://purl.org/orb/Methods)
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-32695266

Paper: Idicula-Thomas S, Gawde U, Bhaye S, Pokar K, Bader GD. Meta-analysis of gene expression profiles of lean and obese PCOS to identify differentially regulated pathways and risk of comorbidities. Comput Struct Biotechnol J 2020. DOI 10.1016/j.csbj.2020.06.023.

Shipped code: https://github.com/bic-nirrh/pcos-metaanalysis @ 204944eb (5 R scripts; no data, no driver, hardcoded G:/, C:«path» paths).

The paper's pipeline (per the Methods + repo scripts)

GEO CEL files (8 datasets)
   │  AffyOldLimma.R (GPL570/GPL96 → affy::justRMA)  |  AffySTLimma.R (GPL6244 → oligo::rma)
   ▼  RMA bg-correct + quantile-norm; probe→gene via BioMart; collapse dup genes by max-SD
limma lmFit + eBayes ModT  →  topTable  →  per-dataset DE table + .rnk (gene, t-value)
   │  significant DEGs: p<0.05 AND (logFC>2 up | logFC<-0.5 down)
   ▼
GSEA v3.0 preranked on .rnk (Bader-lab human GMT)  →  results.edb
   ▼
Cytoscape 3.6.1 Enrichment Map (Cytoscape-EM.R; p=1.0, q=0.1, sim=0.375 COMBINED)
   →  pathway gene-set counts, EM network, top transcription factors
GSVA-HeatMap.R: GSVA of 5 disease gene sets over pooled GPL570 expression → disease/ICD-11 mapping

In scope (attempted) — pipeline-derived, shipped code + public data, deterministic

Result Pipeline Reproducible?
Per-dataset differential expression (RMA + limma ModT) AffyOldLimma.R on GEO CEL files YES — run on GSE98421 + GSE10946(lean)
Significant-DEG identification at paper thresholds (p<0.05, logFC>2 / <-0.5) limma topTable + threshold YES
"7 genes commonly dysregulated across two lean tissues" (PRRT1, SLITRK4, CRHBP, HAPLN1, SRGN, EREG, WNT5A) intersection of per-tissue DEGs YES — adipose (GSE98421) ∩ cumulus (GSE10946-lean), both lean

Chosen datasets: the brief's named accession GSE98421 (lean, subcutaneous adipose, GPL570, 4v4 — unambiguous) plus GSE10946 lean subset (cumulus, GPL570, 6 nonPCOS-Lean vs 5 PCOS-Lean — titles encode group exactly, matches the paper's 6+5). Both run through the authors' AffyOldLimma.R exactly (GPL570 → the script's hardcoded affy_hg_u133_plus_2 BioMart filter is correct here).

Out of scope (not attempted) — and why

Result Why out of scope
Aggregate 5014 DEGs (123 up / 4891 down), 4224 unique, tissue split (1284/2473/202/7/658) Requires running all 8 datasets incl. GPL6244-ST (ambiguous sample selection in GSE98595: 16 samples vs paper's 3v3), GPL15362 (GSE43264, non-Affy custom array — neither shipped script handles it), GPL96 (GSE5090 — script's BioMart filter is hardcoded to U133-Plus-2, wrong for U133A). The last-20% long tail; not pinnable per-dataset (paper gives only the aggregate).
Pathway gene-set counts (86 common / 1031 lean-unique / 159 obese-unique), 4 pathway themes GSEA preranked + Cytoscape 3.6.1 GUI Enrichment Map; needs the Bader-lab GMT (Nov-2018 snapshot) and an interactive desktop app. Non-deterministic to script headlessly.
7 top transcription factors (ETV3, GABPB1, ELF3, GABPA, ELF1, ELF4, SRF) Downstream of the Cytoscape EM network (iRegulon/manual).
GSVA disease enrichment / ICD-11 mapping (222 developmental / 192 metabolic / 127 nervous; 37 visual; 136 PCOS genes) GSVA-HeatMap.R reads unshipped files (Allgenes_3tissues_gpl570.txt, GMT_only5_diseases.txt) at hardcoded G:/ paths — gene sets not provided in the repo.

Verdict framing

This is a partial, faithful reproduction of the differential-expression stage (the deterministic, shipped-code, public-data 80%) on two lean datasets, with an explicit check of the paper's crisp gene-level cross-tissue claim. The downstream GSEA/Cytoscape/GSVA results (the under-specified, GUI-bound, or unshipped-geneset 20%) are deliberately not attempted.

Figures / tables: Table
C1
Reported
7 genes commonly dysregulated in lean PCOS across two tissues: PRRT1, SLITRK4, CRHBP, HAPLN1, SRGN, EREG, WNT5A
Reproduced
All 7 reproduced as significant DEGs (raw p<0.05) in BOTH lean tissues GSE98421 (adipose) and GSE10946-lean (cumulus); cross-tissue intersection = 9 genes containing exactly these 7 (plus IGFBP7, SERPINE2)
exact
C2
Reported
DEG threshold p<0.05 & logFC>2(up)/<-0.5(down); aggregate 5014 DEGs = 123 up / 4891 down (~40:1 down-skew, all 8 datasets)
Reproduced
GSE98421: 152 sig (2 up / 150 down); GSE10946-lean: 908 sig (16 up / 892 down). Strong down-skew faithfully reproduced; aggregate count not directly comparable (only 2 of 8 datasets in scope)
partial
C3
Reported
DEGs called at p<0.05
Reproduced
Confirmed the paper means RAW p (BH-adjusted -> 0 DEGs). Named genes have adj.p ~0.50-1.00 -> do not survive multiple-testing correction in these small studies
within tolerance

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 78/100

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: Q3 · Location of the main deviation 🟡
Minor / cosmetic deviation
+1 pts
From: Q4 · Cause of the deviation 🟡
Concordant (toward reproduced)
Code + data deposited & functional
-2 pts
From: Data & code availability Available & functional
Total score +2

The paper's headline claim — the 7 commonly-dysregulated lean-PCOS genes across two tissues — reproduces exactly (7/7) by running the authors' own shipped AffyOldLimma.R on the public GEO CELs, with no fabrication. Deviations are minor and on the input/our-method side: the intersection is 9 vs 7 (2 extra genes, plausibly BioMart annotation drift) and the aggregate/pathway/comorbidity counts (C4–C7) were out of scope (GUI-bound GSEA/Cytoscape + unshipped gene sets), so they could not be checked. One honest caveat sits on the authors' side: DEGs are nominal raw-p<0.05 and none survive BH correction in these tiny n=4–6/arm studies, a power limitation — not a data mismatch. Overall a high-confidence partial reproduction: 1:1 on the deterministic DE stage, yellow only because much of the paper was not attempted and the raw-p caveat applies.

🤝
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.

194.1 k
tokens (I/O) · 15.9 M incl. cache
23 min
runtime · 0.04 CPU-h
2.5 GB
peak RAM
5 (4 failed)
HPC jobs
hummel
machine