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

Spatial transcriptomics reveals the molecular signatures of prodromal and advanced α-synucleinopathy.

iScience · 2026
L1 70/100 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: 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 +6
✓ What held up
  • Reported values are derivable from the shared data
What did not (or only partly)
  • 🟡Could not use the authors’ exact input data
  • 🟡Reported values were only indirectly comparable
  • 🟡A deviation arose in the data or preprocessing
  • 🟡A deviation was attributed to the published material
  • 🟡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
70/100
Reproducibility score
at the mean
vs. all fields · 1173 studies
🎯 Scores higher than 37% of all assessed papers rank 732 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 for a clean PARTIAL 1:1 of the in-scope target. NOTE the enriched metadata was wrong: the paper states 'No new code was generated' (QuPath is third-party imaging software, not a pipeline/repo), and the primary dataset is GSE274605 (Visium ST), not GSE26927. The paper's primary ST DEG counts (142/132/395/656) were deliberately NOT attempted (TARGET B/C): they depend on manual Loupe Browser per-spot region annotation + ES/LS staging, not a scriptable artifact (the hard 20%). We reproduced TARGET A: the cross-validation of ST signature genes in 4 PUBLIC human PD microarray datasets (GSE26927/GSE7621/GSE20146/GSE43490), by applying the standard curated-GEO limma PD-vs-control DGE workflow the paper itself cites (ref 25) to the paper's own deposited public data (valid per P16). The single most specific, falsifiable claim reproduced EXACTLY: SPP1 up at LogFC>=0.25 in GSE26927/GSE7621/GSE43490 but NOT in GSE20146 (+0.10), and not consistently significant. CREBBP (their protein-validated hit) is up in all 4; NUFIP2 up in 3/4. KCNJ10/GPR37/ROCK2 are strongly up in the SN datasets but reach >=3 only if GSE43490's 3 regions are counted separately (the paper's '>=3 datasets and/or regions' wording) - not chased. Fig S4B inconsistent-direction set: ELAVL4/GABARAPL1/ACHE reproduce as inconsistent; NRXN3/SLC18A2 reproduce as consistently down (canonical PD loss). Overall: described-well-enough, an honest 1:1 with the headline claim exact and the rest partial/qualitative. Fabrication concern: none - all checked values derive from the shipped public data. Not attempted: the GSE274605 ST pipeline and all imaging/wet-lab results.

💻 Code ↗ 🗄 Data: GSE26927

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 70
    assessed: 2026-06-14 ⛓ ba09cb2fe7fb
✎ 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
2026-06-14
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

Does spatial transcriptomics of brainstem α-synuclein pathology in the M83+/+ mouse model reveal stage-specific molecular signatures (prodromal vs. advanced) that are recapitulated in human PD brains and could serve as biomarkers or therapeutic targets?

Core claims
  • Induction of aSyn pathology in rodent brainstem at the early/prodromal stage triggers upregulation of energy metabolism pathways (glycolysis, oxidative phosphorylation, fatty acid metabolism) finding
  • The late/symptomatic stage is characterized by drastic downregulation of mitochondrial metabolic pathways and perturbed mRNA translation machinery, plus inflammatory response finding
  • Aberrant osteopontin (SPP1) signaling and increased CREB-binding protein (CREBBP) expression are consistent markers of progressive aSyn pathology in both the rodent model and PD patient brains finding
  • Spatial transcriptomics applied to sagittal brain sections can spatially resolve gene expression in register with neuroanatomical annotations and disease stage method
  • Cross-validation of rodent ST candidates against 4 independent PD patient microarray datasets identifies conserved molecular signatures method
  • Increased CREBBP expression is a unique marker of advanced α-synuclein pathology finding
  • An online interface/database of the spatial mouse PD transcriptomic data is provided as a community resource resource
  • Aberrant ROCK2 and SPP1 signaling reflect progressive tissue damage mechanism
Experimental setups
Assay System Perturbation Readout Platform
Spatial transcriptomics (ST) M83+/+ (Prnp-SNCA*A53T) transgenic mouse sagittal brain sections intramuscular murine PFF aSyn delivery (early stage DPI-45; late stage DPI-75); PBS control spatially resolved gene expression / differentially expressed transcripts by brain region 10x Genomics Visium v1
Immunofluorescence (IF) M83+/+ mouse brain regions (GRN/pons, PAG/midbrain, DCN/cerebellum, MD/thalamus) intramuscular PFF aSyn vs PBS phosphorylated aSyn (p-aSyn, S129) intensity as % of total area
Immunofluorescence (IF) M83+/+ mouse brain regions (GRN, PAG, DCN, MD) intramuscular PFF aSyn vs PBS CREBBP/CBP intensity as % of cells in total area
Cell-cell communication analysis M83+/+ mouse ST dataset none (computational) ligand-receptor signaling patterns (e.g., SPP1, IGF, FGF) CellChat
Gene set enrichment analysis (GSEA) M83+/+ mouse ST profiles none (computational) HALLMARK and KEGG pathway enrichment / gene ontology MSigDB / STRING
Curated microarray gene expression analysis PD patient brain (SN, LC, GPi/striatum, dmX) from 4 GEO cohorts (GSE26927, GSE7621, GSE20146, GSE43490) PD vs control (none) differential gene expression cross-validating rodent ST candidates GEO microarray datasets
Key results
  • Early stage associated with upregulation of ATP metabolic/biosynthetic processes (glycolysis, OXPHOS, fatty acid metabolism) in disease-affected regions
  • Late stage associated with downregulation of metabolic processes and cytoplasmic translation, plus perturbed microtubule transport, synaptic vesicle exocytosis, neuronal apoptosis, chromatin remodeling
  • Global transcriptomic profiles: ES 961 upregulated / 241 downregulated; LS 669 upregulated / 1,465 downregulated transcripts ES 961up/241down; LS 669up/1465down
  • 16,833 unique protein-coding transcripts identified, clustering by anatomical annotation and disease stage 16,833 transcripts
  • SPP1 (osteopontin) signaling unique to aSyn-pathology cohorts, originating in pons with progressive receptor interactions in midbrain at LS
  • Spp1 among transcripts with global upward trend (up in ES, further increased in LS), along with mt-Nd4l, mt-Atp8, Apod, Rbm3, Gabra1, Adcy1
  • Crebbp/CBP increased abundance in brains of M83+/+ mice as a marker of advanced pathology
  • Late-stage disease-affected regions (pons, midbrain, white matter) enriched for inflammatory response, interferon alpha/gamma, complement, TNF/NF-κB signaling
Key statistics
  • count 16,833 unique protein coding transcripts (total transcripts identified in ST data)
  • count ES: 961 upregulated, 241 downregulated; LS: 669 upregulated, 1,465 downregulated (global DGE profiles by stage)
  • count 1,325 unique transcripts (ES: 142 up, 132 down; LS: 395 up, 656 down) (transcripts prioritized for cross-validation in PD datasets)
  • fold_change Log2 fold change ≥±0.25; adjusted p ≤ 0.05 (cut-off criteria for significant transcripts)
  • count 8 ST annotation/region profiles (brain region annotations delineated)
  • other spot 55 μm diameter; 100 μm center-to-center inter-spot distance (Visium v1 capture resolution)
  • count PBS n=2; ES (DPI-45) n=3; LS (DPI-75) n=3 (cohort sizes for IF; statistical significance not achieved)
  • other 4 independent microarray datasets (GSE26927, GSE7621, GSE20146, GSE43490) (PD patient cohorts for cross-validation)

Statistical methods review

Model: sonnet

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

The study applied 10× Genomics Visium spatial transcriptomics to sagittal brain sections from a transgenic mouse model of α-synucleinopathy at two disease stages (early: n=3, late: n=3, controls: n=2), using a log2 fold-change and adjusted-p-value threshold to identify differentially expressed genes, followed by gene set enrichment analysis (HALLMARK and KEGG via MSigDB) and CellChat ligand-receptor communication analyses. Immunofluorescence protein quantification across the three groups was tested with Kruskal-Wallis ANOVA and Dunn post-hoc comparisons, with results reported as mean ± SD. Candidate transcriptomic signatures from the mouse model were cross-validated by curated analysis of four independent human PD microarray datasets from the GEO repository.

Replicationbiological Sample sizeGroup sizes explicitly stated (PBS n=2, ES n=3, LS n=3); authors acknowledge small cohort size; no formal power calculation described in available text GroupsPBS vehicle control vs. early-stage aSyn pathology (DPI-45) vs. late-stage aSyn pathology (DPI-75) in M83+/+ transgenic mice Pairingunpaired Randomization/blindingnot stated DispersionSD Exact p-valuesyes Effect sizesyes Confidence intervalsno Multiplicity correctionAdjusted p-value for ST DGE (adjustment method not named in available text; implies FDR or similar); Dunn multiple comparisons post-hoc for Kruskal-Wallis immunofluorescence tests
Statistical tests used
Test Applied to n Assumptions
Kruskal-Wallis ANOVA with Dunn post-hoc multiple comparisons Quantification of p-aSyn (S129) immunofluorescence intensity in GRN, PAG, DCN, and MD across three cohorts (Figure 1C) PBS n=2, DPI-45 n=3, DPI-75 n=3 (total n=8) not stated
Kruskal-Wallis ANOVA with Dunn post-hoc multiple comparisons Quantification of CREBBP/CBP immunofluorescence intensity in GRN, PAG, DCN, and MD across three cohorts (Figure 5C) PBS n=2, DPI-45 n=3, DPI-75 n=3 (total n=8) not stated
Differential gene expression analysis with adjusted-p-value cutoff (underlying statistical model not named in available text; STAR Methods not provided) Spatial transcriptomics DGE between PBS, ES, and LS; cutoff: Log2FC ≥ ±0.25 and adjusted p ≤ 0.05 (Tables S2–S5) n=2 PBS, n=3 ES, n=3 LS (sections/animals) not stated
Gene set enrichment analysis (HALLMARK and KEGG gene sets via MSigDB) Pathway-level characterization of global and region-specific ST DGE profiles (Figure 2F, Figure S1B) not stated
CellChat ligand-receptor cell-cell communication analysis Identification of ligand-receptor signaling patterns at ES and LS in ST data (Figure S3) not stated
Curated microarray gene expression analysis (specific test not stated in available text) Cross-validation of mouse ST candidates in four human PD GEO datasets (GSE26927, GSE7621, GSE20146, GSE43490) not stated
Approaches that could also have been used
  • Immunofluorescence group comparisons used Kruskal-Wallis ANOVA with Dunn post-hoc tests, with p-values shown on graphs but all results non-significant
    Could also: Reporting an effect size metric (e.g., eta-squared for Kruskal-Wallis, or rank-biserial correlation for pairwise comparisons) alongside p-values would also characterize the data — With group sizes of n=2–3, no test has meaningful statistical power, making the p-value largely uninformative; an effect size conveys the magnitude of the observed difference independently of sample size and helps readers judge biological relevance
  • Dispersion in immunofluorescence bar graphs was reported as mean ± SD
    Could also: Overlaying individual data points (dot/strip plots) on bar graphs, or replacing bar graphs with beeswarm or jitter plots, would also represent the data at this sample size — With n=2–3 per group, a bar with SD can visually imply more data than exist; showing individual points alongside the mean makes the actual spread and any influential observations fully transparent, which is widely recommended for small-n biological experiments
  • Spatial transcriptomics DGE was performed with a log2FC and adjusted-p threshold, but the underlying statistical model (e.g., negative binomial, linear mixed model) is not named in the available text
    Could also: Explicitly named frameworks such as DESeq2 (negative binomial Wald test), edgeR, NNSVG (spatially aware variance modeling), or SPARK-X could also be applied and reported by name — Naming the DGE model and normalization strategy allows readers to assess its assumptions for count-based ST data and enables full reproducibility; spatially aware methods additionally account for the spatial autocorrelation inherent to Visium data
  • Pathway enrichment was conducted using overrepresentation analysis on a thresholded gene list (Log2FC ≥ ±0.25, adjusted p ≤ 0.05) against HALLMARK and KEGG gene sets
    Could also: Rank-based gene set enrichment analysis (preranked GSEA) using the full continuously ranked gene list (ranked by Log2FC or signed –log10 p-value) could also be applied — GSEA uses the complete ranked gene list rather than a binary threshold, which can detect coordinated pathway-level shifts that do not reach the chosen fold-change or p-value cutoff, and is less sensitive to the choice of thresholds
  • Cross-validation of mouse ST findings in human PD was performed by curated analysis of four independent microarray GEO datasets, examined separately
    Could also: A fixed- or random-effects meta-analysis pooling effect estimates across the four datasets could also be applied to produce a summary effect size with confidence intervals and a heterogeneity estimate (I²) — A meta-analytic approach quantifies the consistency of findings across independent cohorts, distinguishes true cross-dataset replication from dataset-specific signals, and produces an overall effect estimate with formal uncertainty bounds
  • Cell-cell communication analysis was performed with CellChat applied to Visium spatial transcriptomics spot-level data
    Could also: Alternative tools such as NicheNet, LIANA (which aggregates multiple methods), or Squidpy's spatial co-expression modules could also be applied to the same data — Different communication inference tools rely on distinct ligand-receptor databases and scoring algorithms; convergent signals across two or more tools strengthen confidence that identified interactions reflect biology rather than a database or algorithmic artifact
Software: 10× Genomics Visium v1 (ST platform) · CellChat · STRING protein interactions database · MSigDB (HALLMARK and KEGG gene sets) · Gene Expression Omnibus (GEO; for microarray retrieval)

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
0
Impact: low
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.

Jackson Laboratories Cat_004479 RRID in Article (http://semanticscience.org/resource/SIO_001029)
no other assessed paper uses this yet
RRID:AB_143165 RRID in Article (http://semanticscience.org/resource/SIO_001029)
no other assessed paper uses this yet
RRID:AB_3676065 RRID in Article (http://semanticscience.org/resource/SIO_001029)
no other assessed paper uses this yet
RRID:AB_3738383 RRID in Article (http://semanticscience.org/resource/SIO_001029)
no other assessed paper uses this yet
RRID:AB_869973 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-41736854

Title: Spatial transcriptomics reveals the molecular signatures of prodromal and advanced α-synucleinopathy. iScience 2026. DOI 10.1016/j.isci.2026.114845. PMCID PMC12927100.

⚠️ Metadata correction (enriched pointers were wrong/misleading)

The auto-enriched scaffold listed:

  • code_url = https://github.com/qupath/qupath
  • data_accession = GSE26927

Both are misleading once you read the paper's Data and code availability:

Data: … publicly accessible through the NCBI GEO repository (accession number GSE274605). • Code: No new code was generated in the study.

  • QuPath (v0.5.1) is third-party image-analysis software listed in the Key Resources Table, used only for histology/immunofluorescence image quantification (a wet-lab/manual step). It is not an analysis pipeline / authors' repository. → The "code" pointer is a text-mining false positive.
  • GSE274605 is the paper's own primary dataset: Visium spatial transcriptomics (ST) of M83⁺/⁺ mice (Space Ranger 1.3.0 + Seurat 4.3.0).
  • GSE26927 is only one of four pre-existing public PD microarray datasets the authors re-analyzed for cross-species validation. The other three are GSE7621, GSE20146, GSE43490.

There is no authors' code repository. Per brief rule P16, reproducing by applying a standard third-party method to the paper's own data is fully valid.

Reported pipeline-derived results (candidate reproduction targets)

# Result Pipeline Data In scope?
A Cross-validation of ST signatures in human PD microarrays (Fig S4A/B, Table S6): specific genes up/down in PD vs control across 4 GEO microarray datasets Standard curated GEO DGE ("as described previously", ref 25 = Gomes Moreira & Jan, Sci Data 2023) → limma PD-vs-control per dataset 4 public GEO series: GSE26927 (SN), GSE7621 (SN), GSE20146 (GPi), GSE43490 (SN/dmX/LC) YES — primary target (80/20)
B ST DEG counts: 1,325 select transcripts (ES 142↑/132↓; LS 395↑/656↓) Space Ranger 1.3.0 → Seurat SCTransform → FindMarkers (log2FC≥±0.25, adj.p≤0.05); ES/LS staging + region annotation done manually in Loupe Browser GSE274605 NO — hard 20% (see below)
C ST clustering / UMAP / GO / GSVA / KEGG, GO bubble plots Seurat + msigdbr + GSVA + Pathview + STRING GSE274605 NO — depends on B's manual annotation
D Histology: %area pSyn, CREBBP/ROCK2 IF intensity (Fig 6, Fig S6) QuPath image quantification of microscopy microscopy images (not deposited as a pipeline) NO — wet-lab/manual imaging

Why B/C are the hard 20% (deliberately not chased)

The ST pipeline's region labels and ES/LS disease-stage groupings come from a manual step: "spots under tissues were selected in LoupeBrowser, json files were then exported and used for alignment" and regions assigned by hand from a mouse brain atlas. The exact ES-vs-control / LS-vs-ES sample groupings and per-spot annotations driving the 142/132/395/656 counts are not a scriptable artifact, so a faithful 1:1 of the counts isn't cleanly achievable from the deposited data alone. Out of scope per brief rule 3 (do not chase the last 20%).

Primary reproduction target (A) — concrete claims

From the Results section "Curated analyses of patient-derived microarray datasets …" and Fig S4A:

  1. SPP1: increased (LogFC ≥ 0.25) in all microarray datasets except GSE20146 (GPi); pattern not consistently significant (by p-value).
  2. KCNJ10, GPR37, CREBBP, NUFIP2, ROCK2: consistently upregulated in PD vs control across ≥3 datasets/regions.
  3. (Fig S4B, secondary) downregulated-in-ST genes NRXN3, ELAVL4, SLC18A2, GABARAPL1, ACHE: significantly altered but inconsistent direction across the PD microarrays.

These are crisp, falsifiable, gene-level directional claims on fully public data → ideal for an honest 1:1 check.

Met

Figures / tables: Fig S4AFig S4BTableTables
C1_SPP1_up
Reported
SPP1 up (LogFC>=0.25) in all PD microarrays EXCEPT GSE20146; not consistently significant
Reproduced
GSE26927 +0.36, GSE7621 +0.71, GSE43490 +0.73 (all>=0.25); GSE20146 +0.10 (<0.25); adj.p ns in all 4
exact
C2_CREBBP_up
Reported
CREBBP up in >=3 datasets
Reproduced
up>=0.25 in 4/4 (+0.48,+1.60,+0.67,+1.35)
exact
C2_NUFIP2_up
Reported
NUFIP2 up in >=3 datasets
Reproduced
up>=0.25 in 3/4
within tolerance
C2_KCNJ10_up
Reported
KCNJ10 up in >=3 datasets/regions
Reproduced
up>=0.25 in 2/4 (strong in SN datasets)
partial
C2_GPR37_up
Reported
GPR37 up in >=3 datasets/regions
Reproduced
up>=0.25 in 2/4 (strong in SN datasets)
partial
C2_ROCK2_up
Reported
ROCK2 up in >=3 datasets/regions
Reproduced
up>=0.25 in 2/4; positive 3/4
partial
C3_ELAVL4_incons
Reported
ELAVL4 altered, inconsistent direction
Reproduced
-1.02,-1.76,+1.03,+0.13 inconsistent
exact
C3_GABARAPL1_incons
Reported
GABARAPL1 altered, inconsistent direction
Reproduced
-0.36,-0.61,+0.31,-0.07 inconsistent
exact
C3_ACHE_incons
Reported
ACHE altered, inconsistent direction
Reproduced
-0.85,-0.68,-0.30,+0.54 inconsistent
exact
C3_SLC18A2_incons
Reported
SLC18A2 altered, inconsistent direction
Reproduced
consistently DOWN (-2.78,-3.15,-0.67; not on GSE43490)
partial
C3_NRXN3_incons
Reported
NRXN3 altered, inconsistent direction
Reproduced
consistently DOWN (-0.69,-0.96,-0.33; not on GSE43490)
did not match
B_ST_DEG_counts
Reported
1325 ST transcripts (ES 142up/132down; LS 395up/656down)
Reproduced
NOT_ATTEMPTED (out of scope)
partial

Assessments & scoring basis

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

🤖 AI curator · claude (ai-curator room) · v1.0 L1 70/100

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: 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 +6

This is an honest partial reproduction of the secondary target only: the cross-validation of ST signature genes in four public PD microarrays (Fig S4A/B), reproduced via a cited generic limma workflow since the paper deposited no code. The headline falsifiable claim (SPP1 up>=0.25 in every dataset except GSE20146, not consistently significant) reproduced exactly, and protein-validated CREBBP is up 4/4; deviations (C2 genes only 2/4 pooled; NRXN3/SLC18A2 consistently down vs reported 'inconsistent') sit on the our-method/underspecified-wording side, not on the authors' or fabrication side — all checked values derive from the shipped public data. The paper's primary ST conclusion (GSE274605 DEG counts 142/132/395/656) was out of scope (manual Loupe annotation + ES/LS staging), so overall confirmation is solid but limited, and the scaffold metadata (QuPath as code, GSE26927 as primary) is erroneous.

🤝
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.

184.8 k
tokens (I/O) · 12.7 M incl. cache
19 min
runtime · 0.01 CPU-h
1.3 GB
peak RAM
2
HPC jobs
hummel
machine