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

PDGFRA defines the mesenchymal stem cell Kaposi's sarcoma progenitors by enabling KSHV oncogenesis in an angiogenic environment.

PLoS Pathog · 2019
L1 85/100 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 🟡
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 +3
✓ What held up
  • Reported values are derivable from the shared data
  • Any deviation was negligible
  • The central claim held under reproduction
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
  • 🟡Overall, the reproduction showed a material discrepancy
How its reproducibility compares
85/100
Reproducibility score
0.6 SD above mean
vs. all fields · 1173 studies
🎯 Scores higher than 67% of all assessed papers rank 348 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 RNA-seq DEG branch. Two reported host-gene DEG counts were reproduced from the authors' OWN deposited GSE141866 log2-CPM matrix and came out close: 407 vs 454 (~90%) and 1618 vs 1861 (~87%; 1926 at FC>1.5 brackets the reported value). Deliberate, disclosed deviations: (a) paper used DESeq2 on raw TopHat counts, but only a log2-CPM matrix was deposited, so we used limma-trend (the appropriate test for log-CPM); (b) the deposited matrix contains only 2 of the 3 K-Pa(+)S KS-media replicates listed in GEO; (c) nominal p (matches paper phrasing) vs adjusted p. Conclusion: the reported DEG counts are clearly DERIVABLE from the deposited data within ~10-13% despite an engine substitution + a missing replicate -> no fabrication concern. Also corrected a registry error: the harvested accession GSE100684 is a cross-referenced human dataset (PMID 29352292); the study's real data is GSE141868. NOT attempted (80/20): exact DESeq2-on-FASTQ re-run (TopHat 2.1.0 deprecated; raw counts not deposited) and the ChIP-seq branch (kundajelab/AQUAS pipeline, mm9, Drosophila spike-in normalization) - the legitimately heavier ~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.

  1. v1 current initial assessment Score 85
    assessed: 2026-06-15 ⛓ 47e757288677
✎ 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

The authors hypothesize that PDGFRA-positive/SCA-1-positive bone marrow-derived mesenchymal stem cells (Pα(+)S MSCs) are the oncogenic progenitors of Kaposi's sarcoma, and that KSHV induces transformation in these cells via a PDGFRA-driven mechanism only under pro-angiogenic environmental conditions.

Core claims
  • PDGFRA(+)/SCA-1(+) bone marrow-derived MSCs (Pα(+)S MSCs) are KS spindle-cell progenitors finding
  • Pro-angiogenic KS-like environmental conditions are critical/essential for KSHV sarcomagenesis finding
  • Growth in KS-like conditions generates a de-repressed KSHV epigenome that enables oncogenic KSHV gene expression in infected Pα(+)S MSCs mechanism
  • KS-like growth conditions allow KSHV-infected Pα(+)S MSCs to overcome KSHV-driven oncogene-induced senescence and cell cycle arrest via a PDGFRA-signaling mechanism mechanism
  • PDGFRA is both a phenotypic determinant for KS progenitors and a critical enabler of viral oncogenesis mechanism
  • A novel cell-type-defined de novo model of KSHV oncogenesis from primary non-transformed Pα(+)S MSCs resource
  • KSHV establishes stable latent persistent infection in mouse bone marrow-derived MSCs (GFP+/LANA+, no lytic RFP) finding
  • KSHV transcriptomes of MSCs grown in KS-like conditions resemble those of actual human AIDS-KS tumors more than MSC-condition cells finding
Experimental setups
Assay System Perturbation Readout Platform
Flow cytometry / FACS cell sorting Mouse bone marrow-derived MSCs (PDGFRA+/SCA-1+ Pα(+)S and Pα(-)S) none (marker sorting) PDGFRA and SCA-1 expression / cell populations
KSHV infection with fluorescent reporters / fluorescence microscopy Mouse bone marrow-derived Pα(+)S and Pα(-)S MSCs rKSHV.219 infection + puromycin selection GFP (infection) and RFP (PAN promoter lytic) expression rKSHV.219
Immunofluorescence KSHV-infected mouse MSCs (K-Pα(+)S, K-Pα(-)S) and tumors KSHV infection LANA, PECAM1 expression; DAPI nuclei
RT-qPCR K-Pα(-)S and K-Pα(+)S MSCs in MSC vs KS-like media KSHV infection + KS-like media (heparin/ECGF) KSHV gene expression fold-change (LANA, RTA, vGPCR, vIRF1)
Soft agar colony formation assay Pα(+)S KS, K-Pα(+)S MSC, K-Pα(+)S KS cells KSHV infection ± KS-like media Anchorage-independent colony growth
In vivo tumorigenesis (subcutaneous injection) / Kaplan-Meier survival Nude mice injected with Pα(+)S KS, K-Pα(+)S MSC, K-Pα(+)S KS cells KSHV infection + KS-like media Tumor formation / tumor-free survival
Histology (H&E) K-Pα(+)S KS tumor and mECK36 mouse KS-like tumor KSHV infection Tumor histology (vascularized spindle cell sarcoma)
RNA-sequencing (RNA-seq) K-Pα(+)S KS tumors vs in vitro cells; compared to human AIDS-KS biopsies in vivo tumor growth vs in vitro KSHV transcriptome / lytic gene expression; hierarchical clustering
Key results
  • KSHV latent and lytic genes (LANA, RTA, vGPCR, vIRF1) upregulated only in K-Pα(+)S cells in KS-like media
  • Only KSHV-infected PDGFRA-positive MSCs in KS-like conditions (K-Pα(+)S KS) formed colonies in soft agar
  • Subcutaneous injection of K-Pα(+)S KS cells formed tumors in all injected mice by 7 weeks; no tumors from uninfected Pα(+)S KS or infected Pα(+)S in MSC media 6/6 mice
  • KSHV-uninfected/infected PDGFRA-positive or -negative MSCs did not form tumors in nude mice (baseline)
  • KSHV lytic gene expression upregulated in K-Pα(+)S KS tumors in vivo vs tumorigenic cells grown in vitro (in vivo lytic switch)
  • Human KS samples cluster between lytic-expressing mouse KS-like tumors and latently infected K-Pα(+)S KS cells; KS-condition transcriptomes closer to human KS
  • PDGFRA-positive cells were 60% and SCA-1-positive 100% of purified mouse bone marrow-derived MSCs 60% / 100%
  • KSHV de novo infection efficiency similar (~80%) across MSC/KS conditions and PDGFRA-negative/positive cells 80%
Key statistics
  • count 6/6 mice formed tumors by 7 weeks (Subcutaneous injection of K-Pα(+)S KS cells into nude mice (Kaplan-Meier, N=6))
  • count 60% PDGFRA-positive (Fraction of purified mouse bone marrow-derived MSCs positive for PDGFRA)
  • count 100% SCA-1-positive (Fraction of purified mouse bone marrow-derived MSCs positive for SCA-1)
  • count 80% (KSHV de novo infection percentage similar in MSC and KS conditions, PDGFRA-neg and -pos cells)
  • pvalue P < 0.05 (RT-qPCR KSHV gene expression fold-changes in MSC vs KS-like media (triplicates, means ± SD))

Statistical methods review

Model: opus

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 is an experimental cell-biology/virology investigation combining in vitro assays (RT-qPCR gene expression, soft-agar colony formation), in vivo tumorigenesis in nude mice with Kaplan-Meier tumor-free survival, immunofluorescence/histology, and genome-wide RNA-seq with unsupervised hierarchical clustering of KSHV transcriptomes. Quantitative gene-expression results are shown as means of triplicates with SD and significance flagged at a single threshold (*P < 0.05), while survival is summarized by Kaplan-Meier curves with the number of mice (N = 6). Histological tumor assessment was performed by a pathologist in a blinded manner.

Replicationmixed Sample sizeStated as triplicates for RT-qPCR and N = 6 mice for the in vivo survival experiment; no formal power/sample-size calculation described GroupsKSHV-infected vs uninfected PDGFRA+/- MSCs in MSC vs KS-like media; tumor formation across cell/media conditions Pairingunclear Randomization/blindingstated DispersionSD Exact p-valuesno Effect sizesyes Confidence intervalsno Multiplicity correctionnone stated
Statistical tests used
Test Applied to n Assumptions
significance test underlying *P < 0.05 (specific test not named) Fig 1D, fold-changes in KSHV gene expression between 24 hpi and after latency in MSC vs KS-like media (RT-qPCR) triplicates not stated
Kaplan-Meier tumor-free survival (curve shown; comparison test not named) Fig 1F, tumor-free survival after subcutaneous injection into nude mice N = 6 mice na
unsupervised hierarchical clustering Fig 2B, clustering of KSHV transcriptomes from infected cells/tumors and human KS biopsies (RNA-seq) na
Approaches that could also have been used
  • Significance for RT-qPCR triplicates was reported against a single threshold (*P < 0.05) with the specific test unnamed.
    Could also: Naming the exact test (e.g., two-tailed Student's or Welch's t-test, or a nonparametric Mann-Whitney U for small n) and reporting exact P values. — Stating the test and exact P values lets readers see the analysis fully and judge the strength of evidence beyond a pass/fail cutoff.
  • Multiple KSHV genes were compared across conditions, each flagged at P < 0.05.
    Could also: A multiplicity adjustment such as Benjamini-Hochberg FDR or Bonferroni across the family of gene comparisons. — A correction would control the overall false-positive rate when many genes are tested simultaneously.
  • Variability was summarized with SD on n = 3 technical/experimental triplicates.
    Could also: Reporting a 95% confidence interval or showing individual data points alongside the mean. — For small n, plotting individual values and/or a CI conveys both the spread and the precision of the estimate.
  • Kaplan-Meier tumor-free survival curves were presented for the mouse groups.
    Could also: An accompanying log-rank (Mantel-Cox) test or Cox proportional-hazards estimate with a hazard ratio. — A formal survival comparison would quantify the difference between groups and provide an effect estimate with uncertainty.
  • RNA-seq transcriptomes were compared using unsupervised hierarchical clustering.
    Could also: A complementary model-based differential-expression analysis (e.g., DESeq2 or edgeR/limma-voom) with FDR-adjusted results. — A formal differential-expression framework would provide per-gene effect sizes and adjusted significance to support the clustering patterns.
  • Group sizes (triplicates, N = 6 mice) were stated descriptively.
    Could also: An a priori power analysis or rationale for the chosen sample sizes. — Documenting the basis for n helps readers gauge the sensitivity of the experiments to detect the reported effects.

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
33
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.

GQ994935 ENA in Methods (http://purl.org/orb/Methods)
no other assessed paper uses this yet
GSE100684 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-31881074

Paper: Naipauer et al. 2019, PLoS Pathog 15(12):e1008221. "PDGFRA defines the mesenchymal stem cell Kaposi's sarcoma progenitors by enabling KSHV oncogenesis in an angiogenic environment." (mouse Pα(+)S MSC / KSHV model)

Data accession correction (IMPORTANT)

The scaffold/registry recorded data_accession = GSE100684. This is a text-mining false positive. In the paper, GSE100684 is only cross-referenced ("Kaposi's sarcoma KSHV RNA-seq profiles were retrieved from GEO database (GSE100684) from a previous study [43]") — it is human KS data from a different study (PMID 29352292). The authors' own deposited data is GSE141868 (Data Availability statement: "All of the genome-wide data of this study have been deposited in the NCBI Gene Expression Omnibus (GEO) database, GSE number: GSE141868.").

  • GSE141868 SuperSeries (Mus musculus, NextSeq 500), raw reads in SRA SRP237268 / BioProject PRJNA594984:
    • GSE141866RNA-seq (14 samples). Processed file: GSE141866_Preprocessed_datamatrix.csv.gz = log2 CPM matrix.
    • GSE141844ChIP-seq (H3K4me3, H3K27me3).

The repo link github.com/kundajelab/chipseq_pipeline (AQUAS) IS genuinely used by the paper — but for the ChIP-seq branch, not RNA-seq. Per brief P16, applying that third-party tool to the paper's data is a valid reproduction; however it is the heavier ~20% (peak calling, mm9, spike-in normalization) and is secondary.

Pipelines named in Methods (verbatim-sourced from full text)

Result branch Pipeline / tools Reference build
RNA-seq host DEGs TopHat v2.1.0 → count via Rsamtools/GenomicFeatures/GenomicAlignments → DESeq2 GRCm38.82
RNA-seq KSHV transcripts edgeR KSHV 2.0 ref
ChIP-seq host genes AQUAS / kundajelab chipseq_pipeline (BWA 0.7.13, Picard, MACS2) mm9
Functional enrichment ClueGO (Cytoscape), GO/KEGG/Reactome, InnateDB

In-scope (attempted) — RNA-seq DEG counts (low-hanging, clearly specified)

The two reported DEG counts have explicit cutoffs given in the Methods:

  • C1 — 454 DEGs (Fig 4A): K-Pα(+)S KS vs K-Pα(+)S MSC (in vitro, n=3 vs n=3). Cutoff: p-value < 0.01; FC > ±1.5.
  • C2 — 1,861 DEGs (Fig 2F): K-Pα(+)S KS tumors in vivo vs K-Pα(+)S KS cells in vitro (n=8 vs n=3). Cutoff: p-value < 0.001; FC > ±2.

Reproduction approach (honest deviation noted): the authors deposited only the log2-CPM matrix, not raw integer counts; DESeq2 requires raw counts, which were not deposited (only obtainable by re-aligning FASTQ from SRA with the now-deprecated TopHat v2.1.0 — the heavy 20%). We therefore reproduce the DEG counts from the authors' own deposited log2-CPM matrix using a standard, appropriate test for log-CPM (limma moderated t-test) at the paper's exact cutoffs, and report a cutoff-sensitivity sweep so the human auditor sees how recoverable 454 / 1861 are. This is a faithful 1:1 on the deposited intermediate, with the DE-engine substitution (DESeq2→limma) explicitly disclosed.

Out-of-scope (not attempted) — and why

  • Full FASTQ→TopHat→counts→DESeq2 re-run (the exact engine): heavy 20%; TopHat 2.1.0 deprecated; raw counts not deposited. Skipped per 80/20.
  • ChIP-seq AQUAS peak calling (H3K4me3/H3K27me3, mm9, Drosophila spike-in): heavy; spike-in normalization under-specified. Secondary; not attempted unless RNA-seq leaves budget.
  • Wet-lab results (RT-qPCR Fig 1D, flow cytometry, tumor assays): not pipeline.
  • KSHV-transcript edgeR / RPKM tracks: secondary, qualitative figures.

Primary target = C1 + C2 DEG counts from GSE141866 deposited log2-CPM matrix.

Figures / tables: Fig 4AFig 2F
C1
Reported
454 DEGs (K-Pa(+)S KS vs MSC, p<0.01 & FC>1.5, Fig 4A)
Reproduced
407 DEGs (limma-trend on deposited log2-CPM, nominal p)
within tolerance
C2
Reported
1861 DEGs (K-Pa(+)S KS tumors in vivo vs KS cells in vitro, p<0.001 & FC>2, Fig 2F)
Reproduced
1618 DEGs (limma-trend, nominal p); 1926 at FC>1.5
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 85/100

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 🟡
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 +3

The two reported host-gene DEG counts (454 in Fig 4A, 1861 in Fig 2F) were recovered to 407 and 1618 (and 1926 at FC>1.5, bracketing 1861) directly from the authors' own deposited GSE141866 log2-CPM matrix — within ~10-13%, with magnitude and direction intact and no fabrication concern. The deviations sit on our/data-availability side, not the authors': raw TopHat counts were never deposited (forcing a limma-for-DESeq2 substitution) and the matrix holds only 2 of 3 KS replicates. Because the values are clearly derivable and the central transcriptional claim holds, this is a solid 'partial' reproduction whose remaining gaps are explainable methodology/deposit limitations rather than substantive discrepancies.

🤝
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.

170.7 k
tokens (I/O) · 9.5 M incl. cache
16 min
runtime · 0 CPU-h
0.2 GB
peak RAM
1
HPC jobs
hummel
machine