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

Gene-Expression Profiling Suggests Impaired Signaling via the Interferon Pathway in Cstb-/- Microglia.

PLoS One · 2016
54/100 3/4
⚑ Flagged for review — a reproduced result did not match the reported value

Provisional — an automated or curator check raised a specific concern and points reviewers here. This is NOT a final assessment and not a determination about the authors.

Why this verdict

The main results reproduced, with only marginal, non-material deviations.

Reproduced on the brainbox compute brainarbeit.com
How its reproducibility compares
54/100
Reproducibility score
1.1 SD below mean
vs. all fields · 1173 studies
🎯 Scores higher than 14% of all assessed papers rank 997 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

PRELIMINARY (confirmatory re-run in flight). PARTIAL reproduction of the microarray differential-expression arm on the 8 PUBLIC GSE64823 CEL arrays (4 control + 4 Cstb-/-), via a standard third-party pipeline (Bioconductor oligo RMA core -> Welch t-test + BH, limma moderated-t cross-check) on «our HPC» SLURM. The paper's CENTRAL biology reproduces clearly: interferon genes strongly down-regulated in Cstb-/- microglia in both direction and magnitude (Ifit1 -9.94 vs -8.1; Oasl2 -7.63 vs -6.6; Irf7 -5.91 vs -5.3); the knocked-out gene Cstb is itself the single most significant DEG (FC -10.3, adj.p 9.7e-5) -- a clean positive control. The EXACT headline DEG count (155: 151 down/4 up) does NOT reproduce (Welch+BH=2, limma=52), well explained by the public deposit holding 8 arrays vs the paper's 10 (lower power) and open RMA + 20th-pct filter standing in for proprietary Expression Console RMA16 + DABG. The STRT RNA-seq arm (authors' STRTprep tool) is a clean data_unavailable drop -- raw reads never deposited. Provisional grades; human reviewer signs off.

💻 Code ↗ 🗄 Data: GSE64823

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 54
    assessed: 2026-06-21 ⛓ 5eef7a510907
✎ 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
2026-06-24
Rubric version
not recorded
Assessed by
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: sonnet
Founding hypothesis

The paper tests whether loss of cystatin B (CSTB) alters the transcriptional/molecular profile of microglia, to gain insight into the mechanisms underlying the abnormal, dysfunctional microglial activation seen in the Cstb-/- mouse model of EPM1.

Core claims
  • In Cstb-/- microglia, 184 genes were differentially expressed relative to control, of which 33 were identified by both microarray and RNA-seq. finding
  • The vast majority of differentially expressed genes in Cstb-/- microglia are downregulated, not upregulated. finding
  • Several interferon-regulated genes (e.g. Ifit1, Oasl2, Irf7, Isg15, Ifit3) are more weakly expressed in Cstb-/- microglia than in control. finding
  • Downregulation of Irf7 and Stat1 transcripts in Cstb-/- microglia was confirmed by quantitative real-time PCR. finding
  • Functional enrichment and canonical pathway analysis suggest a role for CSTB in chemotaxis, antigen presentation, and immune/defense response processes via altered JAK-STAT pathway signaling. mechanism
  • Gene expression profiles of Cstb-/- microglia are not polarized toward either a pro- or anti-inflammatory status. finding
  • Microarray and RNA-seq expression profiling results for Cstb-/- versus control microglia show good, moderate-to-concordant agreement. finding
  • The findings connect CSTB deficiency in microglia to altered expression of interferon-regulated genes, supporting a role for inflammatory processes in EPM1 disease pathogenesis. mechanism
Experimental setups
Assay System Perturbation Readout Platform
microarray (transcriptome profiling) cultured primary mouse microglia (P5 cortex-derived) Cstb knockout (Cstb-/-) vs wild-type control genome-wide gene expression / differentially expressed genes Affymetrix GeneChip Mouse Exon 1.0 ST array
RNA-seq (STRT protocol) cultured primary mouse microglia Cstb knockout (Cstb-/-) vs wild-type control genome-wide gene expression / differentially expressed genes, spike-in normalized read counts Illumina HiSeq2000
quantitative real-time PCR (qPCR) cultured primary mouse microglia Cstb knockout (Cstb-/-) vs wild-type control relative transcript expression of Stat1, Irf7, Irf9 (Tbp as endogenous control) ABI Prism 7000 Sequence Detection System, TaqMan Gene Expression Assays
JAK-STAT signaling pathway PCR array cultured primary mouse microglia Cstb knockout (Cstb-/-) vs wild-type control expression of JAK-STAT pathway-associated genes Qiagen mouse JAK/STAT Signaling Pathway RT2 Profiler PCR Array, ABI Prism 7000
Gene ontology (GO) enrichment analysis in silico analysis of Cstb-/- microglia DEG list none enriched GO terms among differentially expressed genes GOrilla
canonical pathway / upstream regulator analysis in silico analysis of Cstb-/- microglia DEG list none enriched canonical pathways and predicted upstream regulators (z-score) QIAGEN Ingenuity Pathway Analysis (IPA)
protein-protein interaction network analysis in silico analysis of Cstb-/- microglia DEG list none protein interaction network among differentially expressed genes STRING 10 / Cytoscape 3.2
indirect immunofluorescence cultured primary mouse microglia none culture purity assessment via F4/80 (microglial) and GFAP (astrocytic) marker staining
Key results
  • Microarray identified 155 differentially expressed genes in Cstb-/- microglia: 4 upregulated (FC 1.3 to 1.8, highest Lpcat4) and 151 downregulated (FC -9.3 to -1.3). 151 of 155 DEGs downregulated
  • RNA-seq identified 62 differentially expressed genes in Cstb-/- microglia: 4 upregulated (FC 1.8 to 4.8, highest Serpinb10-ps) and 58 downregulated (FC -200 to -2.2). 58 of 62 DEGs downregulated
  • Interferon-regulated genes Ifit1, Oasl2, and Irf7 were among the most downregulated genes by microarray. Ifit1 FC -8.1; Oasl2 FC -6.6; Irf7 FC -5.3
  • Interferon-regulated genes Oasl2, Isg15, Ifit3, and Ifit1 were among the most downregulated genes by RNA-seq. Oasl2 FC -17.9; Isg15 FC -11.9; Ifit3 FC -11.8; Ifit1 FC -11.2
  • Of 14,420 total genes with expression data, 58% were detected by both microarray and RNA-seq, 20% by RNA-seq only, and 22% by microarray only. 8,378 (58%) shared; 2,844 (20%) RNA-seq only; 3,198 (22%) microarray only
  • Gene expression values correlated moderately between microarray and RNA-seq in both control and Cstb-/- microglia. SCC 0.47 (control), SCC 0.45 (Cstb-/-)
  • Fold changes correlated between microarray and RNA-seq methods. SCC 0.51
Key statistics
  • count 184 differentially expressed genes total (33 identified by both methods) (combined DEGs across microarray and RNA-seq in Cstb-/- microglia)
  • count 155 DEGs (microarray, FC≥1.3, adjusted p<0.05) (microarray differential expression, 5 control vs 5 Cstb-/- pooled samples)
  • count 62 DEGs (RNA-seq, FDR q<0.01, adjusted p<0.01) (RNA-seq differential expression, 5 control vs 5 Cstb-/- samples, 4 technical replicates each)
  • fold_change -8.1 (Ifit1 downregulation, microarray)
  • fold_change -17.9 (Oasl2 downregulation, RNA-seq)
  • correlation SCC: 0.47 (control), 0.45 (Cstb-/-) (Spearman correlation of gene expression values between microarray and RNA-seq)
  • correlation SCC: 0.51 (Spearman correlation of fold changes between microarray and RNA-seq)
  • other RIN > 7.5 (RNA quality of ten control and ten Cstb-/- mice used for extraction)

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 characterized the transcriptome of Cstb-/- vs. wild-type mouse microglia using two parallel platforms: Affymetrix Mouse Exon 1.0 ST microarrays (Welch's t-test + Benjamini-Hochberg correction, fold-change filter) and STRT-based RNA-seq (SAMstrt with ERCC spike-in normalization, FDR + fluctuation thresholds). Platform concordance was assessed by Spearman rank correlation. Selected transcripts were validated by qPCR (Welch's t-test, ddCt method), and functional enrichment was carried out with GOrilla and IPA (Fisher's Exact test). Results were reported primarily as fold changes and threshold p-values.

Replicationmixed Sample sizeTen mice per genotype; cells from 2 mice pooled per RNA sample yielding 5 pooled samples per group; 2 outlier samples removed (1 per group) leaving 4 per group for both platforms; RNA-seq added 4 technical replicates per pooled sample; qPCR used RNA pooled from 2–4 mice per genotype with number of independent pools unstated; no formal power calculation reported GroupsCstb-/- mouse microglia vs. age-matched wild-type microglia (same 129 background) Pairingunpaired Randomization/blindingnot stated Dispersionnone Exact p-valuesno Effect sizesyes Confidence intervalsno Multiplicity correctionBenjamini-Hochberg FDR (microarray); SAMstrt FDR q-value threshold (RNA-seq); no correction stated for qPCR multi-gene testing or GOrilla GO enrichment
Statistical tests used
Test Applied to n Assumptions
Welch's t-test (unpaired, unequal variance) with Benjamini-Hochberg multiple testing correction Microarray differential expression: Cstb-/- vs. control microglia; genes with |FC| ≥ 1.3 and BH-adjusted p < 0.05 considered DEGs 4 pooled RNA samples per group (each pool from 2 mice; 1 outlier array removed per group from original 5) not stated
SAMstrt (spike-in normalized differential expression with FDR and fluctuation testing) RNA-seq differential expression: Cstb-/- vs. control microglia; genes with FDR q < 0.01 and fluctuation adjusted p < 0.01 considered DEGs 4 pooled RNA samples per group × 4 technical replicates each (1 outlier sample removed per group from original 5) not stated
Spearman's rank correlation coefficient Platform comparison of microarray vs. RNA-seq expression values (control SCC 0.47; Cstb-/- SCC 0.45) and fold changes (SCC 0.51) 8378 genes detected by both platforms na
Fisher's Exact test IPA canonical pathway and upstream regulator enrichment analysis; p < 0.05 cutoff; upstream regulators ranked by z-score 184 DEGs (union of microarray and RNA-seq results) not stated
Hypergeometric test (GOrilla) Gene ontology term enrichment of DEGs in Cstb-/- microglia; p < 0.001 considered significant 184 DEGs against expressed background not stated
Unpaired t-test with Welch's correction qPCR validation of Irf7, Stat1, and Irf9 transcript levels (ddCt method, TBP endogenous control); p < 0.05 RNA pooled from 2–4 mice per genotype; number of independent pools not stated not stated
Approaches that could also have been used
  • RNA from two mice of the same genotype was pooled into a single sample before RNA extraction, yielding 5 (then 4) pooled biological units per group
    Could also: Individual per-animal RNA samples could have been profiled separately without pooling, using a linear mixed model or limma/voom to model inter-animal variance — Individual samples preserve inter-animal biological variance, provide more degrees of freedom for statistical testing, and allow detection of outlier animals; pooling conflates biological variability and reduces the effective number of independent replicates
  • Microarray DEGs were defined by a combined hard fold-change cutoff (|FC| ≥ 1.3) and BH-adjusted p < 0.05
    Could also: A ranked-gene approach such as Gene Set Enrichment Analysis (GSEA) or a moderated t-statistic (limma) without a fold-change gate could also be applied — An arbitrary fold-change threshold can exclude precisely measured small-effect genes; rank-based methods evaluate the full expression distribution continuously and are less sensitive to threshold choice
  • Three qPCR target genes (Irf7, Stat1, Irf9) were each tested individually with Welch's t-test at α = 0.05 with no stated correction for the multiple comparisons
    Could also: A Bonferroni correction, Benjamini-Hochberg FDR, or a repeated-measures ANOVA with planned contrasts across the three genes would also control the family-wise error rate — Testing three correlated hypotheses simultaneously at α = 0.05 nominally inflates the type I error rate; a correction maintains the declared error probability across the gene panel
  • Platform concordance was summarized with Spearman rank correlation (SCC ≈ 0.45–0.51)
    Could also: A Bland-Altman (limits-of-agreement) plot or Passing-Bablok regression could also characterize platform agreement — Correlation measures monotonic association but not absolute agreement; Bland-Altman analysis quantifies systematic bias and the spread of differences between platforms, which is typically more informative for method-comparison studies
  • Dispersion of expression values is not reported for the primary biological comparisons; results are given as fold changes and threshold p-values
    Could also: Reporting SD, SEM, or 95% CI alongside fold changes, or showing individual data points per sample, would also convey within-group variability — Dispersion measures allow readers to judge effect consistency across the small number of samples (n = 4 per group) and are particularly informative when sample sizes are low
  • Two outlier arrays (one per genotype) were identified by PCA and hard-excluded, reducing each group from 5 to 4 samples
    Could also: A sensitivity analysis comparing results with and without the excluded samples, or a robust weighting approach (e.g., Cook's distance filtering in DESeq2 or iterative outlier down-weighting in limma), could also be applied — Hard exclusion without sensitivity analysis makes it difficult to assess whether conclusions are driven by the removal decision; robust methods can down-weight influential observations while retaining all data
Software: GeneSpring GX 12 · GraphPad Prism 6.02 · DataAssist (Applied Biosystems) 3.01 · STRTprep / SAMstrt pipeline · TopHat · Casava (Illumina) 1.8.2 · GOrilla (web-based GO enrichment tool) accessed 22.07.2015 · IPA (QIAGEN Ingenuity Pathway Analysis) accessed 22.07.2015 · STRING (protein interaction database) 10 · Cytoscape 3.2

What was reproduced

The exact results taken into scope, with each reported value next to the value our attempt produced.

Scope — pmid-27355630

Paper: Körber et al. 2016, Gene-Expression Profiling Suggests Impaired Signaling via the Interferon Pathway in Cstb−/− Microglia. PLoS One 11(6):e0158195. PMID 27355630 / PMC4927094.

Named code: https://github.com/shka/STRTprep (the authors' own STRT RNA-seq preprocessing+DE pipeline; shka = co-author Shintaro Katayama). HEAD commit 6b033186cff838db29849c78580201c677ff39ec ("Update after the publication", 2016-03-24), no release tags on the cloned tree. Named data: GEO GSE64823.

The paper has TWO independent transcriptomic arms

Arm Platform Analysis software Data deposited?
Microarray Affymetrix Mouse Exon 1.0 ST RMA16 (Affy Expression Console) + DE at FC≥1.3, p<0.05 (BH) YES → GEO GSE64823 (public GSE64823_RAW.tar, CEL files)
RNA-seq STRT single-cell tagged RT, HiSeq2000 STRTprep (bowtie1+TopHat, mm9, SAMstrt DE) NO public accession (verified below)

In scope vs out of scope

IN SCOPE — Microarray differential-expression pipeline (REPRODUCED)

The microarray CEL files are public (GSE64823). The analysis is a standard, well-specified bioinformatic pipeline: RMA normalization → per-gene fold change

  • moderated t-test → threshold |FC|≥1.3 & p<0.05 (BH) → DEG list. Per BRIEF HARD RULE 2 / P16, running a standard third-party R pipeline (Bioconductor oligo RMA + limma) on the paper's own data is an equally valid reproduction. Target claims: # DEGs (155: 151 down / 4 up), strong downregulation of interferon genes (Ifit1, Oasl2, Irf7), direction of effect. Pipeline: oligo::rma() (core transcript clusters) + limma + BH FDR.

OUT OF SCOPE / DROP — STRTprep RNA-seq pipeline (data_unavailable)

STRTprep operates exclusively on raw STRT FASTQ reads. The paper deposits only the microarray data; no public accession exists for the STRT RNA-seq raw reads. Verified 2026-06-16:

  • PMC/PLoS Data Availability statement names only GEO GSE64823 (microarray).
  • ENA/BioStudies: only BioProject PRJNA271951 = the GSE64823 microarray series; no SRA/ENA study holds the STRT reads.
  • GSE64823 platform = GPL6096 (Exon array) only — no sequencing samples. Therefore STRTprep cannot be exercised on this paper's data. This arm is a clean data_unavailable drop (not an env/effort failure). Its RNA-seq DE numbers (62 DEGs etc.) are recorded in claims.tsv as reported-but-not-reproducible.

80/20 decision

Reproduce the microarray DEG pipeline (low-hanging, public data, standard tool). Do NOT attempt: (a) STRTprep RNA-seq (no data); (b) exact RMA16/Expression-Console "expressed transcript cluster" filter (14 147) which depends on a DABG detection step in proprietary Affymetrix software — approximated, flagged as method-dependent; (c) qPCR/JAK-STAT array validation (wet-lab, out of scope).

MA-KO-positive-control
Reported
Cstb knocked out
Reproduced
Cstb is #1 DEG, FC -10.3, adj.p 9.7e-05
exact
MA-IFN-Ifit1
Reported
-8.1
Reproduced
-9.94
within tolerance
MA-IFN-Oasl2
Reported
-6.6
Reproduced
-7.63
within tolerance
MA-IFN-Irf7
Reported
-5.3
Reproduced
-5.91
within tolerance
MA-DEG-total
Reported
155 (151 down/4 up)
Reproduced
2 (Welch+BH) / 52 limma (51 down/1 up)
did not match
MA-DEG-down
Reported
151
Reproduced
51 (limma)
partial
MA-DEG-up
Reported
4
Reproduced
1 (limma); Lpcat4 +1.85
did not match
MA-expressed-clusters
Reported
14147
Reproduced
18665 (20th-pct approx of DABG)
partial
MA-samples
Reported
5+5 (10)
Reproduced
4+4 (8 deposited)
did not match
RS-DEG-total
Reported
62 (58 down/4 up)
Reproduced
not reproducible (no public raw reads)
m.public.grade.error

Assessments & scoring basis

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

No assessment has been recorded yet.
🤝
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.

641.6 k
tokens (I/O) · 45.7 M incl. cache
628 min
runtime
Per-job HPC accounting not captured for this run — the runtime shown is the reproduction’s measured wall-clock time.