Identification of intrinsic genes across general hypertension, hypertension with left ventricular remodeling, and uncontrolled hypertension.
The main results reproduced, with only marginal, non-material deviations.
Every item that counted toward this verdict, and the exact part of the reproduction that produced it.
- ✓Same input data as the authors
- ✓Reported values were directly comparable
- ✓No authors-side cause for any deviation
- ✓Any deviation was negligible
- 🟡A deviation arose in the data or preprocessing
- 🟡Reported values were not (fully) derivable from the shared data
- 🟡The central claim did not (fully) hold under reproduction
- 🟡Overall, the reproduction showed a material discrepancy
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 PARTIALLY reproduce, and the core pipeline reproduces cleanly 1:1. The brief's code link (github.com/cran/RVA) is a third-party DEG-VISUALIZATION helper, not the authors' analysis code; the actual pipeline (limma -> fgsea Hallmark -> WGCNA -> Metascape) is described in Methods and I ran it on the paper's own four public GEO microarray series (GSE24752/75360/74144/71994) via «our HPC» SLURM jobs on «infra» (conda R 4.1.3 + GEOquery/limma/fgsea/msigdbr). RESULT: the two headline DEG counts reproduce within 1-3 genes -- GSE24752 77 vs reported 75, GSE74144 HT-vs-control 25 vs reported 23 (both within-tol); the two 'few DEGs' datasets reproduce as ~0. A consistent up/down LABEL inversion across both datasets indicates a fold-change reference-direction convention (paper defines FC vs the disease group), NOT a discrepancy -- flipping signs matches the paper's split (66/11 ~ 63/12; 3/22 ~ 2/21). No fabrication indicator. GSEA stretch: 6 of 9 paper-named Hallmark pathways recover as significant (FDR<0.05,|NES|>1) in the correct dataset, HT-LVR a clean 3/3 (IFN a/g, glycolysis, DNA repair); the remaining 3 (KRAS, mitotic spindle, coagulation) appear with the expected sign below FDR. NOT attempted: WGCNA module-trait correlations (C5-C7, parameter-sensitive, deprioritised); Metascape GO/KEGG/TF + MCODE hub genes + Cytoscape (web-service/GUI, not deterministically scriptable); and the final intrinsic-gene lists (C8) which derive from an under-specified union/intersection of DEGs+WGCNA+Metascape and are not directly derivable from the text (flagged unverifiable, not asserted wrong). Minor version drift: limma 3.50.3 vs 3.48.3, msigdbr 7.5.1 vs Hallmark v7.4. Verdict: a genuine, well-described study whose core computational outputs reproduce faithfully; partial because the harder WGCNA/Metascape/meta-analysis layers were not (or could not be deterministically) rerun.
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.
-
v1 current initial assessment Score 73assessed: 2026-06-16 ⛓ 695170f63c0a
✎ 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.
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
- 2026-06-16
- Rubric version
- v1.0
- Assessed by
-
🤖 AI curator · claude (ai-curator room) · v1.0 · run #1 2026-06-16no 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: sonnetThe study tests whether distinct sets of 'intrinsic genes' and biological pathways can be identified that characterize different stages of hypertension progression—general hypertension, hypertension with left ventricular remodeling (HT-LVR), and uncontrolled hypertension (UN-HT)—using integrated bioinformatics analysis of peripheral blood transcriptomic datasets.
- ★ FBXW4 and 13 other genes are uniquely enriched in the general hypertension group finding
- ★ TRIM11 and 40 other genes are mainly involved in the hypertension with left ventricular remodeling group finding
- ★ F13A1 and 17 other genes are significantly enriched in the uncontrolled hypertension group finding
- ★ The precise switch of the 'immune-metabolic-inflammatory' loop pathway is the most significant hallmark across different stages of hypertension progression mechanism
- ★ Heme metabolism, TNF alpha/NFkB, interferon alpha response signaling, and MYC target v1/v2 pathways are enriched at different hypertension stages finding
- WGCNA combined with GSEA better identifies functional gene modules than standard DEG thresholding when few DEGs are detected method
- Peripheral blood is a suitable surrogate tissue for discovering hypertension-related genes and pathways resource
| Assay | System | Perturbation | Readout | Platform |
|---|---|---|---|---|
| microarray gene expression profiling / DEG analysis | human peripheral blood cells (GSE24752) | hypertension vs normotensive | differentially expressed genes (fold change, p-value) | Affymetrix Human Genome U133 Plus 2.0 Array (GPL570) |
| microarray gene expression profiling / DEG analysis | human peripheral blood mononuclear cells (GSE75360) | hypertension vs normotensive | differentially expressed genes; WGCNA module-trait correlation | Illumina HumanHT-12 v.4.0 Expression BeadChip (GPL10558) |
| microarray gene expression profiling / DEG analysis | human white blood cells (GSE74144) | hypertension with left ventricular remodeling vs normal LV vs control | differentially expressed genes; WGCNA module-trait correlation | Agilent-026652 Whole Human Genome Microarray 4x44K v2 (GPL13497) |
| microarray gene expression profiling / DEG analysis | human peripheral blood mononuclear cells (GSE71994) | uncontrolled vs controlled hypertension | differentially expressed genes; WGCNA module-trait correlation | Affymetrix Human Gene 1.0 ST Array (GPL6244) |
| gene set enrichment analysis (GSEA) | all four datasets (peripheral blood) | none (in silico pathway analysis) | normalized enrichment score (NES) and FDR for MSigDB hallmark gene sets | R package fgsea v1.18.0 |
| weighted gene co-expression network analysis (WGCNA) | GSE75360, GSE74144, GSE71994 peripheral blood datasets | none (in silico network analysis) | co-expression modules and module-trait correlation coefficients | R package WGCNA v1.70-3 |
| functional enrichment analysis (GO/KEGG/transcription factor enrichment) | WGCNA gene modules of interest | none (in silico) | enriched biological terms and regulatory transcription factors | Metascape |
| multi-gene list meta-analysis / PPI network construction with MCODE | gene sets across HT, HT-LVR, UN-HT | none (in silico) | intrinsic/hub genes and network interactions | Metascape and Cytoscape v3.8.2 |
- – 75 DEGs identified in GSE24752 (63 up, 12 down) 63 up / 12 down
- ▼ 23 DEGs identified in GSE74144 (2 up, 21 down) 2 up / 21 down
- – Few DEGs found in GSE75360, GSE71994, and part of GSE74144 under absolute 2-fold change, p<0.05 threshold
- ▲ SkyBlue module strongly correlated with hypertension (GSE75360) r=0.56, p=0.008
- ▲ Cyan module strongly correlated with hypertension with left ventricular remodeling (GSE74144) r=0.39, p=0.02
- ▲ GreenYellow module most associated with uncontrolled hypertension (GSE71994) r=0.35, p=0.03
- – HEME_METABOLISM, KRAS_SIGNALING_UP, MITOTIC_SPINDLE enriched in hypertensive group; MYC_TARGETS_V1, OXIDATIVE_PHOSPHORYLATION enriched in normotensive group
- ▲ TNFA_SIGNALING_VIA_NFKB, IL2_STAT5_SIGNALING, INTERFERON_ALPHA/GAMMA_RESPONSE enriched in uncontrolled hypertension and HT-LVR groups
- correlation r = 0.56, p = 0.008 (SkyBlue module correlation with hypertension trait, GSE75360 (soft threshold beta = 8))
- correlation r = 0.39, p = 0.02 (Cyan module correlation with hypertension with left ventricular remodeling, GSE74144 (soft threshold beta = 10))
- correlation r = 0.35, p = 0.03 (GreenYellow module correlation with uncontrolled hypertension, GSE71994 (soft threshold beta = 12))
- count 75 DEGs (63 up, 12 down) (GSE24752 hypertensive (n=3) vs normotensive (n=3))
- count 23 DEGs (2 up, 21 down) (GSE74144 hypertensive with normal LV (n=14) vs control (n=8))
- fold_change absolute 2-fold change, p<0.05 (DEG significance threshold used across all datasets)
- count FBXW4 + 13 other genes unique to hypertension group (Multi-gene list meta-analysis intrinsic gene identification)
- count TRIM11 + 40 other genes mainly involved in HT-LVR group (Multi-gene list meta-analysis intrinsic gene identification)
Statistical methods review
Model: sonnetA neutral, descriptive read of the statistical approach — what was done, and (for shared learning, not as criticism) what could also have been done.
This bioinformatics re-analysis study downloaded four public GEO microarray datasets (GSE24752, GSE75360, GSE74144, GSE71994) representing three hypertension stages and applied limma-based moderated linear models to identify DEGs, fgsea-based GSEA for pathway enrichment, and WGCNA to detect trait-correlated co-expression modules. Results across datasets were integrated using Metascape's multi-gene-list meta-analysis with PPI network construction to identify stage-specific intrinsic gene sets; no primary data were collected, and all inference was drawn from secondary analysis of publicly archived expression matrices.
| Test | Applied to | n | Assumptions |
|---|---|---|---|
| limma moderated linear model (empirical Bayes t-statistic) | DEG identification in all six pairwise comparisons across GSE24752, GSE75360, GSE74144, GSE71994 | varies by dataset: 3 vs 3 (GSE24752); 10 vs 11 (GSE75360); 14 vs 8 and 14 vs 14 (GSE74144); 20 vs 20 (GSE71994) | not stated |
| fgsea adaptive Monte Carlo permutation-based GSEA (pre-ranked, nperm=10000) | Pathway enrichment across all six comparisons using MSigDB Hallmark gene sets (H collection, 50 gene sets) | same sample sizes as DEG comparisons above | not stated |
| Pearson correlation (module-trait relationship in WGCNA) | Correlation of module eigengenes with hypertension-stage traits in GSE75360, GSE74144, GSE71994 | GSE75360 n=21; GSE74144 n=36; GSE71994 n=40 | not stated |
| Hypergeometric/Fisher enrichment test (Metascape GO/KEGG) | Functional enrichment of WGCNA modules of interest and multi-gene-list meta-analysis | null | not stated |
| MCODE graph-clustering algorithm | Hub gene identification within PPI network from multi-gene-list meta-analysis | null | na |
-
DEGs were filtered using a raw p-value threshold of < 0.05 combined with an absolute fold-change ≥ 2, applied across thousands of probes simultaneously↳ Could also: An adjusted p-value (e.g., Benjamini-Hochberg FDR) could also be used as the primary significance threshold for DEG calling — When testing tens of thousands of probes simultaneously, unadjusted p-values yield an expected number of false positives proportional to the number of tests; BH-FDR is the standard approach in limma workflows and would explicitly control the expected proportion of false discoveries, making the DEG list more conservative but more reproducible
-
The four datasets were analyzed independently for DEGs and enrichment, then integrated post-hoc using Metascape multi-gene-list meta-analysis↳ Could also: Formal statistical meta-analysis methods such as RankProd, Fisher's combined p-value, or integrated analysis via MetaDE or inSilicoMerging could also aggregate evidence across datasets before downstream analysis — Formal meta-analysis pools per-gene evidence from all datasets simultaneously, increasing statistical power to detect consistently dysregulated genes and providing a single ranked gene list with quantified cross-study consistency, rather than relying on post-hoc overlap of separately derived gene lists
-
GSEA was performed using only the MSigDB Hallmark collection (50 curated gene sets)↳ Could also: Broader MSigDB collections such as C2 (curated pathways: KEGG, Reactome, WikiPathways) or C5 (GO terms) could also be used — The Hallmark collection summarizes well-defined biological states but is deliberately compact; C2 and C5 collections offer finer-grained pathway resolution and may capture disease-relevant biology not represented in the 50 Hallmark gene sets, at the cost of a larger multiple-testing burden
-
Hub genes within the PPI network were identified using the MCODE graph-clustering algorithm via Metascape↳ Could also: Network centrality measures — degree centrality, betweenness centrality, or PageRank — applied to the full PPI graph could also identify hub nodes — MCODE identifies densely connected subgraphs (complexes), while centrality measures identify individual nodes that act as topological bottlenecks or highly connected hubs; both approaches are standard and complementary, and reporting both can distinguish module-level from node-level importance
-
WGCNA module selection relied primarily on the highest Pearson correlation coefficient between the module eigengene and the trait of interest↳ Could also: Module membership (kME, intramodular connectivity) combined with gene significance (GS, correlation of individual gene expression with the trait) could also be used to refine gene selection within the chosen module — Using both kME and GS as quantitative filters within the top-correlated module is a standard WGCNA practice that restricts downstream analysis to genes most robustly associated with the trait, reducing noise from genes that happen to cluster together for unrelated reasons
-
Datasets were analyzed on their original platforms separately without batch correction or cross-dataset normalization, and integration was performed at the gene-list level↳ Could also: Cross-platform harmonization methods such as ComBat, SVA (surrogate variable analysis), or COCONUT could also be applied before pooling samples for joint analysis — Batch effects between platforms (Affymetrix U133, Illumina HT-12, Agilent 4x44K) can inflate or mask true biological signal; explicit batch correction prior to joint analysis is an alternative that permits direct statistical comparison across datasets, though it requires careful evaluation of confounding between batch and biological group
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.
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.
What was reproduced
The exact results taken into scope, with each reported value next to the value our attempt produced.
Assessments & scoring basis
Each contributor’s verdict, the per-question basis, and the auditable, itemised worksheet behind it.
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.
Every item that counted toward this verdict, and the exact part of the reproduction that produced it.
The core computational pipeline reproduces faithfully: DEG counts land within 1–3 genes (GSE24752 77 vs 75; GSE74144 25 vs 23), the two 'few DEGs' datasets give ~0, and 6/9 named Hallmark pathways recover as significant in the correct dataset (HT-LVR a clean 3/3). The only deviations are an internally consistent fold-change direction convention (label inversion, same magnitudes), a minor GSE71994 cohort split (17/23 vs 20/20), and MSigDB/limma version drift — all on our/technical side, with no fabrication signal. However, the paper's titular result — the intrinsic-gene lists (C8) plus the WGCNA modules (C5–C7) — relies on an under-specified DEG+WGCNA+Metascape meta-analysis and is not directly derivable from the shared data, so the central conclusion is only partially confirmed. Overall: solid reproduction of the supporting layers with explainable deviations, but the headline claim remains unverified rather than refuted.
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-4-8Measured resources invested to assess this paper — sanitised (machine class only, no job ids/paths). Compute = HPC accounting (SLURM); tokens = the AI agent's session.