Microglial Fkbp5 Impairs Post-Stroke Vascular Integrity and Regeneration by Promoting Yap1-Mediated Glycolysis and Oxidative Phosphorylation.
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.
- Nothing in this column.
- 🟡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
- 🟡Reported values were not (fully) derivable from the shared data
- 🟡The deviation was non-trivial in magnitude
- 🟡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
PARTIAL reproduction, described well enough to run independently. The paper is a wet-lab/transgenic-mouse mechanism study whose single-cell bioinformatics is a RE-ANALYSIS of public datasets with standard Seurat/Monocle/CellChat - there is NO authors' own analysis-code repo (the registry's DNBelab/MGI code_url is a text-mining false positive; reproduced with Seurat per brief P16). The registry's pinned data accession GSE233812 is only the external-validation set (and belongs to PMID 39499634); the paper's PRIMARY analysis and the headline Fkbp5 number come from GSE174574 (3 sham + 3 MCAO mouse brain scRNA-seq), so we reproduced on GSE174574. Standard Seurat 5.3.0 re-analysis on «our HPC» («job»; 58,332 cells post-QC, 26 clusters, 11 canonical cell types, microglia n=13,163 balanced 6,544 sham/6,619 MCAO): the central QUALITATIVE claim reproduces 1:1 and independently - Fkbp5 is significantly UP-regulated in stroke microglia (avg_log2FC +0.33, p_adj=5.8e-27, detection 11.6%->20.2%). The exact reported magnitude |log2FC|=2.56 was NOT reproduced in bulk microglia: it is specific to the paper's bespoke stroke-VAM subcluster, which I deliberately did NOT reconstruct (the hard ~20%; cluster IDs are resolution-dependent and not 1:1). Honest auditable note (not fabrication): in this primary dataset a Fkbp5 |log2FC|~2.5 actually appears in endothelial/oligodendrocyte/astrocyte cells, not microglia, so the microglia-specificity + 2.56 magnitude should be human-checked against whether the paper means bulk microglia or the stroke-VAM subset. NOT attempted (out of scope): authors' own undeposited snRNA-seq, the 12 microglial subclusters/stroke-VAM=cluster6 numbering, ssGSEA M2 ranking, Monocle pseudotime, CellChat, Hippo/Yap1 target quantitation, phosphoproteomics, and all wet-lab/imaging/behavioural results.
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 50assessed: 2026-06-16 ⛓ f96f79372f51
✎ 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: opusThe paper investigates the unclear role of microglia in blood–brain barrier (BBB) leakage and neovascularization after ischemic stroke, hypothesizing that a distinct perivascular microglial niche (stroke-VAM) regulated by Fkbp5 drives BBB disruption and impaired vascular regeneration.
- ★ A post-stroke perivascular microglial niche (stroke-VAM) exists, characterized by low M2 marker expression and elevated glycolysis, OXPHOS, and phagocytic activity. finding
- ★ Fkbp5 acts as a central regulator driving BBB disruption and impaired neovascularization through stroke-VAM. mechanism
- ★ Fkbp5 inhibits Yap1 phosphorylation, facilitating its nuclear translocation via the Hippo signaling pathway. mechanism
- ★ Microglial Fkbp5 conditional knockout (Cx3cr1Cre Fkbp5flox/flox) increases M2 marker expression, reduces glycolysis/OXPHOS/phagocytosis, decreases BBB leakage, and enhances angiogenesis after stroke. finding
- ★ Fkbp5 cKO enhances interactions between stroke-VAM and endothelial cells, affecting signaling that maintains BBB integrity and promotes neovascularization. finding
- ★ Fkbp5 is upregulated specifically in microglia after stroke, peaking at day 1, and is induced by blood-derived macromolecules (fibrinogen, albumin, serum) in a dose-dependent manner. finding
- Microglial Fkbp5 expression positively correlates with vascular (fibrinogen) leakage and with glycolysis and OXPHOS scores but not TCA. finding
- snRNA-seq, flow cytometry, and Seahorse metabolic assays in transgenic mice were used to characterize Fkbp5-mediated stroke-VAM. method
| Assay | System | Perturbation | Readout | Platform |
|---|---|---|---|---|
| single-cell RNA sequencing (scRNA-seq, public dataset reanalysis) | ipsilateral hemisphere of tMCAO and sham mice, 1 day post-stroke | tMCAO (ischemic stroke) | microglial heterogeneity, cluster identity, ssGSEA metabolic scores | GSE174574 / GSE233812 / GSE225948 datasets |
| single-nucleus RNA sequencing (snRNA-seq) | Cx3cr1Cre Fkbp5flox/flox (Fkbp5 cKO) mice, ipsilateral hemisphere | Fkbp5 conditional knockout in microglia + tMCAO | cell-cell interactions, Hippo pathway alterations, transcriptome | — |
| immunofluorescence staining | peri-infarct brain sections of tMCAO and sham mice | tMCAO | colocalization of Iba1/Hk2/CD31, Iba1/Atp5a1/CD31, Fkbp5/Iba1/CD31/fibrinogen | — |
| qRT-PCR | ipsilateral cerebral cortex; BV2 cells | tMCAO time course (1,3,5,7 days); fibrinogen/albumin/serum treatment | Fkbp5 mRNA level | — |
| Western blot / immunoblotting | ischemic cortex; BV2 cells | tMCAO time course; OGD/R model | Fkbp5 protein expression | — |
| Seahorse metabolic assay | microglia / transgenic mice | Fkbp5 modulation | glycolysis and OXPHOS (metabolic flux) | — |
| flow cytometry | transgenic mice (Fkbp5 cKO) | Fkbp5 cKO + ischemic stroke | microglial phenotype markers | — |
| in vitro OGD/R (oxygen-glucose deprivation/reperfusion) model | BV2 microglial cell line | OGD/R | Fkbp5 protein levels and fluorescence intensity | — |
- ▲ Fkbp5 exhibited the most significant change among candidate regulators in stroke-VAM (absolute Log2FC = 2.56). Log2FC = 2.56
- ▲ Fkbp5 mRNA peaked at day 1 post-stroke and declined to baseline by day 5; protein elevated at days 1 and 3.
- ▲ The number of stroke-VAM among Fkbp5-positive microglia was nearly double that in non-perivascular microglia. ~2-fold
- ▲ Microglial Fkbp5 levels positively correlated with fibrinogen leakage levels.
- ▲ Fkbp5 mRNA induced by fibrinogen, albumin, and serum in a dose-dependent manner in BV2 cells.
- ▲ Clusters 6 and 11 (stroke-VAM) showed elevated glycolysis and OXPHOS ssGSEA scores in tMCAO vs sham, with low M2 scores (ranked 12th and 9th).
- ▲ Perivascular microglia showed significantly increased Hk2 and Atp5a1 expression (colocalized with Iba1 near CD31) in tMCAO vs sham.
- – Fkbp5 expression positively correlated with glycolysis and OXPHOS scores but not TCA score.
- fold_change absolute Log2FC = 2.56 (Fkbp5 change in stroke-VAM)
- count ~double (stroke-VAM in Fkbp5+ vs non-perivascular microglia) (perivascular vs non-perivascular Fkbp5 microglia quantification, n=8)
- correlation positive correlation (microglial Fkbp5 vs fibrinogen leakage) (vascular leakage association)
- count n = 7 (Hk2/Iba1/CD31 colocalization, tMCAO and sham) (Figure 1L–N immunostaining quantification)
- count n = 8 (Atp5a1/Iba1/CD31 colocalization) (Figure 1O–Q immunostaining quantification)
- count n = 9 (Iba1/CD31 colocalization, tMCAO and sham) (Figure 1R quantification)
- count n = 4 per group (qPCR Fkbp5 time course) (Fkbp5 mRNA across tMCAO time points)
- count n = 8 per group (Fkbp5 protein immunoblot time course) (Figure 2E,F Western blot quantification)
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.
The study combined reanalysis of publicly available scRNA-seq datasets (GSE174574, GSE233812, GSE225948) with in vivo mouse tMCAO experiments. Single-cell analyses used ssGSEA scoring, unsupervised clustering, pseudotime, and GO enrichment to characterize microglial heterogeneity. Experimental validation (immunofluorescence, qRT-PCR, western blot) used unpaired t-tests for two-group comparisons and one-way ANOVA for multi-timepoint comparisons, with results reported as mean ± SD and significance denoted by threshold-based asterisks.
| Test | Applied to | n | Assumptions |
|---|---|---|---|
| Unpaired t-test | Two-group immunofluorescence quantification: tMCAO vs sham for Hk2/Iba1, Atp5a1/Iba1, and Iba1/CD31 colocalization; perivascular vs non-perivascular microglial Fkbp5 expression | n = 7–9 per group | not stated |
| One-way ANOVA | Multi-timepoint comparisons of Fkbp5 mRNA (qPCR) and protein (western blot) at sham, 1, 3, 5, 7 days post-tMCAO; multi-group immunofluorescence panels | n = 4 (qPCR) or n = 8 (western blot, IF) per group | not stated |
| Single-sample gene set enrichment analysis (ssGSEA) | Scoring of M1, M2, DAM, glycolysis, OXPHOS, TCA, phagocytosis, and mitochondrial gene programs across 12 microglial subclusters in scRNA-seq data (GSE174574); compared between tMCAO and sham within each cluster | null | na |
| Differential expression analysis (specific method not stated) | Identification of subcluster-specific marker genes across 12 microglial subclusters; DEGs stratified by Fkbp5-high vs Fkbp5-low expression; multi-criteria Venn overlap to screen stroke-VAM regulators | null | not stated |
| Gene Ontology (GO-BP) enrichment analysis | Cluster-specific DEGs for microglial subclusters 6 and 11; Fkbp5-high and Fkbp5-low DEG sets; threshold P < 0.05 | null | not stated |
| Correlation analysis (method not specified) | Microglial Fkbp5 expression vs fibrinogen leakage levels; Fkbp5 expression vs glycolysis, OXPHOS, and TCA ssGSEA scores | null | not stated |
-
One-way ANOVA was used for multi-group/multi-timepoint comparisons, but the post-hoc correction procedure was not named↳ Could also: Specifying a post-hoc test — e.g., Tukey HSD for all pairwise comparisons or Dunnett's test when comparing all time points to a single sham control — would also address these multi-group comparisons — Named post-hoc procedures make the family-wise error control strategy explicit and reproducible; this is standard practice in multi-group animal experiments and aids readers in assessing which specific contrasts drove significance
-
Dispersion was reported as mean ± SD for experimental groups with n = 4–9 animals↳ Could also: Individual data points overlaid on bar or dot plots, or 95% confidence intervals, could also convey the data distribution — With small n, showing individual points alongside the mean makes the underlying variability and any outliers directly visible; CI-based reporting also communicates precision of the estimate, and both approaches are increasingly recommended by journals for small-sample preclinical studies
-
Differential expression analysis was performed on scRNA-seq microglial subclusters but the specific statistical method (e.g., Wilcoxon rank-sum, MAST, DESeq2) was not stated in the available text↳ Could also: Pseudo-bulk approaches (e.g., aggregating cells per biological sample, then applying DESeq2 or edgeR) could also be used when multiple biological replicates contribute cells to each cluster — Pseudo-bulk methods account for within-sample correlation among cells and avoid the inflated effective sample sizes that arise when each cell is treated as an independent observation; reporting the chosen method aids reproducibility
-
ssGSEA scores were compared visually between tMCAO and sham groups within each microglial subcluster (scatter/violin plots), without a stated formal inferential test on those scores↳ Could also: A Wilcoxon rank-sum or t-test on per-cell ssGSEA scores between conditions, with FDR adjustment across gene sets, could also formally quantify and correct for these multi-set comparisons — Formalizing the comparison with a test and correction makes the evidence for metabolic shifts in stroke-VAM quantitatively reproducible and guards against inflated type-I error when many gene sets are scored simultaneously
-
Significance was communicated exclusively through threshold-based asterisk notation rather than exact p-values↳ Could also: Reporting exact p-values alongside a standardized effect-size metric (e.g., Cohen's d for t-tests, eta-squared for ANOVA) would also be standard — Exact p-values allow readers to judge the strength of evidence without the binary threshold framing, and effect sizes contextualize whether statistically significant differences are also practically meaningful — particularly relevant for small-n animal studies
-
The correlation between microglial Fkbp5 expression and fibrinogen leakage was reported as 'significant positive correlation,' but the correlation method (Pearson vs Spearman) was not specified↳ Could also: Spearman's rank correlation could also be used, especially if immunofluorescence intensity data are skewed or contain outliers; reporting the correlation coefficient and its 95% CI would further quantify the association — Specifying the method and assumptions (normality, linearity) clarifies the robustness of the reported association and allows direct comparison with correlation data from other 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.
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-41355597
Paper: Li Y et al. Microglial Fkbp5 Impairs Post-Stroke Vascular Integrity and Regeneration by Promoting Yap1-Mediated Glycolysis and Oxidative Phosphorylation. Adv Sci (Weinh) 2025. PMID 41355597 · PMCID PMC13042415 · DOI 10.1002/advs.202512499.
Nature of the paper's computation
This is primarily a wet-lab / transgenic-mouse mechanism paper. Its single-cell bioinformatics is a re-analysis of THREE public scRNA-seq datasets plus the authors' own (apparently undeposited) snRNA-seq of Fkbp5-conditional mice. The tools are standard: Seurat (clustering/UMAP/DE), Monocle (pseudotime), CellChat (cell–cell communication), KEGG/GO/ssGSEA.
Note on the registry's pinned "code" + "data"
- code_url = DNBelab_C_Series_scRNA-analysis-software (MGI): this is almost certainly a text-mining false positive. The analyzed datasets are 10x/Illumina (GPL21103/GPL24247/...), not MGI DNBelab data, and the paper's methods name Seurat/Monocle/CellChat, not DNBelab. There is no authors' own analysis-code repository. Per brief rule P16, this is fine: we reproduce the pipeline-derived result by running a standard third-party tool (Seurat) on the paper's own (public) data, which is equally valid.
- data_accession = GSE233812: this is only the paper's external validation dataset (Fig S3R–T; it actually belongs to a different study, Zucha et al. PMID 39499634). The paper's PRIMARY re-analysis (Fig 1–2, the Fkbp5 screen) is GSE174574 (Zheng/Hao, scRNA-seq MCAO vs sham 24 h, 3+3 samples, 10x). We therefore reproduce on GSE174574, the dataset that actually yields the headline number.
Public datasets the paper re-analyzes
| GEO | role in paper | figure |
|---|---|---|
| GSE174574 | PRIMARY: microglial heterogeneity + stroke-VAM + Fkbp5 screen | Fig 1, Fig 2 |
| GSE233812 | external validation of stroke-VAM existence | Fig S3R–T |
| GSE225948 | microglia-specific Fkbp5 expression | Fig S4E |
IN SCOPE (pipeline-derived, attempted) — on GSE174574
- C1 (structural): "16 cell clusters annotated by canonical markers" (Fig 1A–B). Reproduce a standard Seurat clustering of the 6 samples and report the major cell types recovered + cluster count. (Resolution-dependent → expect partial.)
- C2 (HEADLINE): Fkbp5 is a top DEG up-regulated in stroke microglia, reported |log2FC| = 2.56 in stroke-VAM (Fig 2; text). Reproduce by computing Fkbp5 differential expression in the microglia compartment, MCAO vs sham. Check direction (up), significance, and magnitude vs 2.56.
- C3 (supporting): Fkbp5 induction is microglia-specific (vs other major cell types) within the same dataset (supports Fig S4E claim).
OUT OF SCOPE (not attempted — the hard ~20% or non-pipeline)
- Authors' own snRNA-seq of Fkbp5 ΔMG / fl-fl mice (7754 / 5812 cells, 10 clusters, MG1–MG6): own data, not deposited → unobtainable.
- Exact 12 microglial subclusters + "stroke-VAM = cluster 6" numbering: bespoke resolution/integration choices, not 1:1 reproducible (cluster IDs are arbitrary).
- ssGSEA M2 "ranked 12th/9th", per-subcluster glycolysis/OXPHOS module scores, Monocle pseudotime, CellChat Wnt/NRG/PDGF, Hippo/Yap1 target quantitation, phosphoproteomics (Lats1 S278/S871): wet-lab or bespoke-pipeline 20%, skipped.
- All immunostaining, behavioural, vascular, biochemical results: wet-lab, non-pipeline, out of scope.
Pipeline named per in-scope result
All of C1–C3: Seurat standard workflow (QC → LogNormalize → HVG → PCA →
neighbors/Louvain clustering → UMAP → canonical-marker annotation →
FindMarkers Wilcoxon for Fkbp5 MCAO-vs-sham within microglia). Data fetched
INSIDE the «our HPC» compute job to «infra»; no data on «host».
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 is a wet-lab mechanism study whose scRNA-seq is a re-analysis of public data; reproduction on the paper's actual primary dataset (GSE174574, not the wrongly-pinned GSE233812) confirms the core qualitative claim — Fkbp5 is significantly up-regulated in stroke microglia (+0.33 log2FC, p_adj=5.8e-27, detection 11.6%→20.2%). The deviations are mostly on our methodology side (the stroke-VAM subcluster carrying the reported |log2FC|=2.56 was deliberately not reconstructed, so bulk microglia is diluted) and on data-availability (microglia-specificity rests on a separate dataset GSE225948). One auditable discrepancy, not fabrication: in the primary dataset a ~2.5 Fkbp5 effect sits in endothelial/oligodendrocyte/astrocyte cells rather than microglia, so a human should verify whether the 2.56 is correctly attributed to (sub)microglia. Overall solid-partial with explainable, non-suspicious deviations.
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.