Comprehensive data for studying serum exosome microRNA transcriptome in Parkinson's disease patients.
The main results reproduced: recomputed values matched the published ones within tolerance.
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 relevant deviation in data/preprocessing
- ✓No authors-side cause for any deviation
- ✓Reported values are derivable from the shared data
- ✓Any deviation was negligible
- ✓The central claim held under reproduction
- ✓Overall, the reproduction was clean
- Every checked point held up.
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 -> EXACT 1:1 reproduction. Sci Data data descriptor (Yu et al. 2024); analysis repo SSaikiLab/SciData (renamed AMED-Rasagiline/SciData, commit 455af86) ships, per 4 cohort-years, a DESeq2 R script + its input count matrix + its expected output CSVs. Reran the shipped DESeq2 pipeline (filter rowMedians>0, half-min zero-impute, DESeq2 two-group/before-after design, padj<=0.05 DEG calls) on the shipped count matrices (R 4.3.3, DESeq2 1.42.0, «our HPC» «job»). All 13 quantitative claims reproduced exactly: results-table row counts (775/433/798/623), increased-miRNA counts (196/69/202/67), decreased-miRNA counts (160/183/152/179); per-miRNA log2FoldChange Pearson r=1.0 with max abs diff <=3e-9 (float noise), and every Increased/Decreased DEG set identical (Jaccard=1.0, zero set differences). Deviations were numerics-preserving only: fixed one hard-coded input path (2020); omitted the 2022 read of Clinical-Parameters.xlsx (absent from repo, but its covariates are not in the ~Time design so cannot affect results -- the exact match confirms this); stripped interactive plot/Shiny calls that emit no numbers. No fabrication concern: every shipped value is regenerable from shipped data+code. NOT attempted (out of scope): upstream FASTQ->count-matrix small-RNA-seq quantification (raw reads not shipped), wet-lab/clinical/neuroimaging tables (non-computational), and the downstream GSEA ranking xlsx (optional last 20%).
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 100assessed: 2026-06-14 ⛓ 848d85fc3723
✎ 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
- 2026-06-14
- Rubric version
- v1.0
- Assessed by
-
🤖 AI curator · claude (ai-curator room) · v1.0 · run #1 2026-06-15no 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: sonnetInvestigating the association between serum exosome miRNA transcriptome levels and demographic, clinical, and neuroimaging characteristics could expand the landscape of biomarkers and therapeutic targets in Parkinson's disease, and rasagiline-treatment-induced changes in the miRNA transcriptome could highlight key pathogenetic factors.
- ★ The study presents comprehensive serum exosome miRNA transcriptome data from four independent Japanese cohorts of PD patients and controls. resource
- ★ Two PD/Control cohorts comprise 96 PD patients and 80 age- and gender-matched controls with anonymised demographic, clinical, and neuroimaging data for PD patients. resource
- ★ Two rasagiline cohorts comprise 96 PD patients profiled before and after one year of rasagiline treatment. resource
- ★ Serum exosome miRNA profiles from PD patients and healthy individuals cluster into two distinct groups by PCA, indicating a global effect of PD on the miRNA transcriptome. finding
- ★ miRNAs upregulated in PD patients are significantly enriched in categories linked to neurodegenerative diseases, inflammation, and exosome localisation. finding
- ★ miR-550a-3p levels are consistently positively associated with the MIBG heart/mediastinum ratio in PD patients across two cohorts. finding
- ★ Serum exosome miRNA changes in this dataset significantly overlap with independently published PD (CD4+ T-cell) and Alzheimer's disease (brain) miRNA datasets, supporting data reliability. finding
- ★ Ten candidate miRNAs were identified as increased in PD in the 2020 and 2021 cohorts and decreased following rasagiline treatment in the 2022 cohort. finding
| Assay | System | Perturbation | Readout | Platform |
|---|---|---|---|---|
| Small RNA sequencing (miRNA transcriptome) | Serum exosomes, PD patients vs age/gender-matched controls (2020, 2021 cohorts) | none | miRNA expression levels | Ion Total RNA-Seq Kit v2 on Ion PGM System (Thermo Fisher Scientific) |
| Small RNA sequencing (miRNA transcriptome) | Serum exosomes, PD patients before/after treatment (2022, 2023 cohorts) | drug (rasagiline, 1 year) | miRNA expression levels | Ion Total RNA-Seq Kit v2 on Ion PGM System (Thermo Fisher Scientific) |
| RNA/miRNA quality control | Serum exosome extracts from PD patients and controls | none | Total RNA and miRNA quantity/quality (electropherograms) | Nanodrop One; Agilent Total RNA 6000 Pico Kit and Small RNA Kit on Agilent 2100 Bioanalyzer |
| 123I-MIBG scintigraphy | PD patients (cardiac sympathetic nerves) | none | Heart/mediastinum (H/M) and thyroid/mediastinum (T/M) ratios | E CAM scintigraphy camera with smartMIBG software |
| DAT-SPECT imaging | PD patients (striatal dopaminergic terminals) | none | Specific Binding Ratio (SBR) | Siemens Symbia T16 with DAT VIEW software |
| Principal component analysis (bioinformatics) | Serum exosome miRNA profiles, 2020 cohort | none | Sample clustering by miRNA expression | pcaExplorer (R package) |
| miRNA gene set enrichment / overrepresentation analysis | Serum exosome miRNAs upregulated in PD | none | Enriched disease, inflammation, and localisation categories | miEAA |
| Correlation analysis | Serum exosome miRNA (miR-550a-3p) vs MIBG H/M ratio, PD patients | none | Spearman correlation coefficient | — |
- – PCA clustered serum exosome miRNA profiles of PD patients and controls into two distinct groups in the 2020 cohort.
- ▲ Upregulated miRNAs in PD significantly enriched in 'neurodegenerative diseases' category. padj=0.001
- ▲ Upregulated miRNAs in PD significantly enriched in 'inflammation' category. padj=0.007
- ▲ Upregulated miRNAs significantly enriched in the 'exosome' localisation term. padj=2.44×10^-5
- ▲ miRNAs increased in PD (2020 cohort) overlapped significantly with miRNAs increased in CD4+ T-cells of PD patients. padj=0.0004
- ▼ Downregulated miRNAs in PD (2020 cohort) overlapped significantly with miRNAs downregulated in Alzheimer's disease brain. padj=3×10^-11
- ▲ miR-550a-3p positively correlated with MIBG heart/mediastinum ratio in the 2020 cohort. spearman r=0.4106, p=0.0218
- ▲ miR-550a-3p positively correlated with MIBG heart/mediastinum ratio in the 2021 cohort. spearman r=0.3489, p=0.0466
- correlation spearman's r=0.4106, p=0.0218 (miR-550a-3p vs MIBG H/M ratio, 2020 cohort)
- correlation spearman's r=0.3489, p=0.0466 (miR-550a-3p vs MIBG H/M ratio, 2021 cohort)
- pvalue padj=0.001 (enrichment of upregulated PD miRNAs in 'neurodegenerative diseases' category)
- pvalue padj=0.007 (enrichment of upregulated PD miRNAs in 'inflammation' category)
- pvalue padj=2.44×10^-5 (enrichment of upregulated PD miRNAs in 'exosome' localisation term)
- pvalue padj=0.0004 (overlap of PD-increased serum exosome miRNAs with CD4+ T-cell PD miRNAs)
- pvalue padj=3×10^-11 (overlap of PD-decreased serum exosome miRNAs with Alzheimer's disease brain miRNAs)
- count 96 PD patients and 80 controls (PD/Control cohorts); 96 PD patients (rasagiline cohorts) (cohort sample sizes across the four studies)
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 data descriptor paper presents four cohort datasets of serum exosome miRNA transcriptomes in Parkinson's disease (PD). Statistical analyses were used primarily for technical validation rather than primary hypothesis testing: principal component analysis (PCA) visualised global miRNA profile differences between PD patients and controls, miRNA enrichment analyses (GSEA and overrepresentation) assessed biological relevance of differentially expressed miRNAs, and Spearman's correlations examined associations between individual miRNA levels and neuroimaging measures. Adjusted p-values (padj) were reported for enrichment results; exact p-values and Spearman's r were reported for correlation analyses.
| Test | Applied to | n | Assumptions |
|---|---|---|---|
| Principal component analysis (PCA) via pcaExplorer R package | Visualisation of global serum exosome miRNA profiles in PD patients vs. healthy controls (2020 cohort) | 46 PD patients + unnamed number of controls from 2020 cohort | not stated |
| miRNA gene set enrichment analysis (GSEA) via miEAA, ranked by FC × −log10(padj) | Enrichment of PD-vs-control differentially expressed miRNAs in disease and localisation categories | — | not stated |
| Gene overrepresentation analysis via miEAA | Overrepresentation of upregulated miRNAs in brain disorder and inflammation categories | — | not stated |
| Study/study overlap comparison (hypergeometric-type; padj reported) via Makjaroen et al. database | Overlap of differentially expressed miRNAs in 2020 cohort with external PD CD4+ T-cell and Alzheimer's brain datasets | — | not stated |
| Spearman's rank correlation | Correlation between miR-550a-3p expression and MIBG heart/mediastinum ratio in 2020 and 2021 PD/Control cohorts | 2020: n=46 PD; 2021: n=50 PD (inferred from Table 1) | not stated |
| Differential miRNA expression analysis (specific test not stated in text; code available on GitHub; performed in R 4.2.2) | Identification of miRNAs significantly increased in PD vs. controls and decreased following rasagiline treatment | 96 PD + 80 controls (PD/Control cohorts); 96 PD before/after rasagiline (rasagiline cohorts) | not stated |
-
Spearman's correlation was used to relate miR-550a-3p to the MIBG heart/mediastinum ratio across two cohorts, with each tested separately and unadjusted p-values reported↳ Could also: A combined analysis with Bonferroni or FDR correction across the two correlation tests, or a meta-analytic combination of the two Spearman's r estimates, could also be applied — Pooling or formally correcting across two tests on the same hypothesis would quantify cumulative evidence and account for the multiplicity of testing the same miRNA-neuroimaging pair in two cohorts
-
Demographic continuous variables (age, H&Y stage, UPDRS-III) were summarised with Mean(SD)↳ Could also: Median and IQR, or reporting both Mean(SD) and range, would also be standard for clinical data that may not be normally distributed — Clinical rating scales such as H&Y stages are ordinal; non-parametric summary statistics and distributional visualisations often complement mean-based summaries for such variables
-
PCA was used to visualise global miRNA expression differences between PD patients and controls↳ Could also: UMAP or t-SNE dimensionality reduction could also have been applied for high-dimensional miRNA count data — UMAP and t-SNE can better preserve local cluster structure in high-dimensional omics data and may reveal subgroupings not apparent in the first two principal components
-
Differential miRNA expression between groups was assessed (specific test not stated in text) using R, with results characterised as 'significantly increased' or 'decreased'↳ Could also: DESeq2 or edgeR (negative binomial GLM-based tests) are widely used for count-based RNA-seq differential expression and could also be applied to miRNA count matrices — Count-based RNA-seq models explicitly account for overdispersion and library-size variation common in small RNA sequencing data, and have established normalisation procedures for cross-sample comparison
-
The four cohorts (2020, 2021, 2022, 2023) were analysed and reported largely as independent datasets, with cross-cohort agreement assessed by Venn diagram overlap↳ Could also: A formal meta-analysis or mixed-effects model pooling data across cohorts, or a replication-based framework treating one cohort as discovery and another as validation, could also be used — Quantitative pooling or a staged discovery-replication design would yield a single combined effect estimate with confidence intervals and formally test whether findings replicate across cohorts
-
miEAA enrichment results were reported with padj values and visualised categorically, without reporting enrichment scores or effect sizes for individual enriched categories↳ Could also: Normalised enrichment scores (NES) from GSEA, or odds ratios with confidence intervals from overrepresentation analysis, could also be reported alongside adjusted p-values — Effect-size metrics such as NES or OR convey the magnitude of enrichment, not just its statistical significance, which aids interpretation and comparison across studies
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.
Assessed papers, coloured by verdict. Click a node to open it.
- 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.
Scope — pmid-39406833
Paper: Yu et al. 2024, Scientific Data — "Comprehensive data for studying serum exosome microRNA transcriptome in Parkinson's disease patients." PMID 39406833 · PMCID PMC11480472 · DOI 10.1038/s41597-024-03909-6
Type: Data Descriptor. Primary contribution = the dataset (GEO GSE269781, super-series of GSE269775/6/7/9). The repo ships a preliminary DESeq2 analysis of the four cohort count matrices as the computational/technical-validation layer.
Code: github.com/SSaikiLab/SciData → renamed to AMED-Rasagiline/SciData
(GitHub 301). Pinned commit 455af865a1c5b6ca004486d16e27444b51392b98
(default branch main, pushed 2024-09-09). License: none declared.
In scope — pipeline-derived results (DESeq2 differential miRNA)
The repo ships, for each of 4 cohorts, a *.R script + the input count matrix +
the expected output CSVs. The pipeline per cohort is identical and deterministic:
- Load count matrix (xlsx/csv).
- Coerce to integer; keep miRNAs with
rowMedians > 0(detected in ≥ half). - Impute remaining zeros as
round((rowMin+1)/2)(half-min). DESeq2::DESeqDataSetFromMatrix+DESeqwith a two-level design, thenresults().- Increased =
log2FC>0 & padj<=0.05; Decreased =log2FC<0 & padj<=0.05. resultstable written withRank = log2FC * (-log10 padj)(2020/21) /* abs(log10 padj)(22/23).
DESeq2 has no stochastic component → expect (near-)exact reproduction.
| cohort | GEO sub-series | design | input file | expected outputs (ground truth) |
|---|---|---|---|---|
| 2020 (PD vs HC) | GSE269775 | ~Disease, 50 PD / 50 HC | 2020/GSE269775_2020_CountMatrix.xlsx | results.csv (775), Increased (196), Decreased (160) |
| 2021 (PD vs HC) | GSE269776 | ~Disease, 46 PD / 30 HC | 2021/2021.xlsx | results.csv (432), Increased (69), Decreased (183) |
| 2022 (before/after rasagiline) | GSE269777 | ~Time, 46×(Before,After) − PD10 | 2022/AMED-2022-...csv | GSEA-Raw-Data (798), Increased (202), Decreased (152) |
| 2023 (before/after rasagiline) | GSE269779 | ~Time, 50 Before / 50 After | 2023/AMED-2023-...xlsx | GSEA-Raw-Data (623), Increased (67), Decreased (179) |
Out of scope (not attempted)
- Upstream FASTQ→count-matrix small-RNA-seq alignment/quantification: raw reads not in repo; count matrices are the shipped starting point. (Pipeline named in GEO, not rerun.)
- Wet-lab (exosome isolation, library prep), clinical/neuroimaging tables — not computational.
- Interactive/visual steps in the scripts (
plotPCA,pcaExplorerShiny app,EnhancedVolcano) — produce no numeric output; stripped so the job is non-interactive. - GSEA ranking (
miRNA-GSEA-Results.xlsx) downstream of the ranked list — optional 20%.
Faithful-rerun deviations (documented, do not change numerics)
- Path fix only: 2020.R hard-codes
Documents/AMED-.../...xlsx; pointed at the in-repo copy. Other scripts use repo-relative paths already. - 2022: script reads
Clinical-Parameters.xlsx, which is not in the repo. Those covariates (Therapy/Gender/HY/Age) are not in the design formula (~Time), so they cannot affectresults. The covariate-read block is omitted; the~TimeDESeq2 is run verbatim. Flagged in AUDIT.md. - Removed interactive/plot calls (see out-of-scope) — numeric DESeq2 path untouched.
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.
This Scientific Data descriptor ships, per cohort-year, a DESeq2 R script + input count matrix + expected output CSVs; rerunning the shipped pipeline on the shipped matrices reproduced all 13 quantitative claims exactly (775/433/798/623 results rows; 196/69/202/67 increased; 160/183/152/179 decreased), with per-miRNA log2FC Pearson r=1.0 and identical DEG set membership (Jaccard=1.0). Deviations were purely numerics-preserving (a hard-coded path fix, omission of a covariate read absent from the ~Time design, and stripped interactive plotting calls). Residual differences are pure float noise (≤3e-9 in log2FC), which is the strongest possible evidence against fabrication. The reproduction is on our/technical side only and severity is negligible — a clean 1:1.
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.