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

Downregulation of Splicing Factor PTBP1 Curtails FBXO5 Expression to Promote Cellular Senescence in Lung Adenocarcinoma.

Curr Issues Mol Biol · 2024
L1 59/100 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
Scoring basis — itemised

Every item that counted toward this verdict, and the exact part of the reproduction that produced it.

Main result did not reproduce
Decisive
From: Q5 · Derivability / plausibility 🔴
✓ What held up
  • Nothing in this column.
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
  • 🔴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
How its reproducibility compares
59/100
Reproducibility score
0.9 SD below mean
vs. all fields · 1173 studies
🎯 Scores higher than 19% of all assessed papers rank 925 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? PARTIALLY. The handed dataset GSE63577 is real, public, and ships processed count/RPKM matrices, so the 80% (the senescence DE signature) was reproducible on «our HPC» with DESeq2 1.50.2 WITHOUT re-running STAR on 48 raw libraries. RESULT IS MIXED: the paper's BIOLOGY reproduces cleanly (positive control LMNB1 strongly down; FBXO5 AND PTBP1 both robustly downregulated in senescence across all 5 fibroblast lines, significant in 4/5) -> 1:1 on the qualitative premise. But the paper's HEADLINE NUMBER, '355 universally underexpressed genes', is NOT reproducible: the paper specifies no threshold, no statistical test, and no definition of 'universally', and is internally inconsistent about whether it spans 3 or 5 cell lines. A strict significant-DE intersection across the 3 stated lines (MRC5/IMR90/WI38) yields 0 genes (IMR90 has only 1 significant DE gene due to pathological dispersion); method-dependent variants span 238-4937 with no natural criterion hitting 355. Flagged as possible-non-derivable (number not regenerable from shipped data as described). NOT ATTEMPTED (the hard 20%): (C2) 756 rMATS splicing events - the A549 PTBP1-KD RNA-seq has no public accession in the paper, so it is not reproducible from public data; (C3) 2276 TCGA-LUAD DEGs - needs TCGA download plus an unspecified PTBP1 high/low grouping threshold. STAR (the harvested 'code' link) is a third-party tool; per BRIEF rule 2 we validly reproduced by applying a standard DE pipeline to the paper's own data.

💻 Code ↗ 🗄 Data: GSE63577

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 59
    assessed: 2026-06-14 ⛓ b454fbecab70
✎ 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

Whether regulation of the splicing factor PTBP1 influences cellular senescence in lung adenocarcinoma (LUAD), and how PTBP1-mediated alternative splicing of downstream targets such as FBXO5 controls senescence.

Core claims
  • PTBP1 is overexpressed in LUAD and associated with poor prognosis/tumor growth, indicating an oncogenic role. finding
  • PTBP1 expression is decreased in multiple cellular senescence models, opposite to its trend in cancer. finding
  • Suppression of PTBP1 induces cellular senescence and G2/M cell cycle arrest in LUAD cells. finding
  • PTBP1 silencing enhances exon 3 skipping of FBXO5, generating the less stable FBXO5-S splice variant and reducing overall FBXO5 expression. mechanism
  • Downregulation of FBXO5 induces senescence and G2/M arrest in LUAD cells. finding
  • FBXO5-S has a faster mRNA degradation rate than FBXO5-L, explaining reduced FBXO5 levels. mechanism
  • Targeting PTBP1-mediated aberrant splicing has therapeutic potential for senescence-mediated tumor suppression in cancer. finding
  • RNA-seq with rMATS identified PTBP1-dependent alternative splicing events, intersected with cell-cycle genes to nominate FBXO5. method
Experimental setups
Assay System Perturbation Readout Platform
bulk RNA-seq (with rMATS alternative splicing and DESeq2 DEG analysis) PTBP1-knockdown A549 LUAD cells vs control shRNA knockdown of PTBP1 alternative splicing events (SE/MXE/A3SS/A5SS/IR) and differentially expressed genes Illumina HiSeq; STAR, StringTie, rMATS, DESeq2 v1.26.0
RT-qPCR / RT-PCR A549 and H1299 LUAD cell lines shRNA knockdown of PTBP1 or FBXO5 mRNA levels of PTBP1, FBXO5, and FBXO5-S/FBXO5-L isoforms (normalized to GAPDH) LightCycler 480; SYBR qPCR Master Mix (EZBioscience)
Western blot A549 and H1299 LUAD cell lines shRNA knockdown of PTBP1 or FBXO5 PTBP1 and FBXO5 protein levels (GAPDH control) anti-PTBP1 (Sangon D225103), anti-FBXO5 (Proteintech 10872-1-AP); ECL Bio-Rad
SA-β-Gal staining A549 and H1299 LUAD cell lines shRNA knockdown of PTBP1 or FBXO5 senescence-associated β-galactosidase activity Beyotime C0602 kit
Cell cycle flow cytometry A549 and H1299 LUAD cell lines shRNA knockdown of PTBP1 or FBXO5 DNA content / cell cycle distribution (G1, G2/M) BD Biosciences flow cytometer; PI/RNase staining; ModFit 3.0
mRNA stability assay (Actinomycin D chase) A549 and H1299 LUAD cell lines ActD 10 μg/mL transcription block (0–8 h) decay rate of FBXO5-S vs FBXO5-L by RT-qPCR
CCK-8 proliferation, colony formation, and transwell migration assays A549 and H1299 LUAD cell lines shRNA knockdown of PTBP1 proliferation rate, colony number, migrated cell number CCK-8 (Dojindo); Corning transwell 8 μm; ImageJ v2.3.0
Bioinformatic database / public dataset analysis TIMER (33 cancers), GEPIA (542 LUAD vs 59 normal), TCGA-LUAD, GEO GSE63577 senescence models (BJ, IMR90, WI38, HFF, MRC5) none PTBP1/FBXO5 expression, survival, DEGs, pathway enrichment (KEGG/GO via DAVID)
Key results
  • PTBP1 is significantly increased in LUAD and other tumors vs normal tissue
  • Higher PTBP1 levels associated with lower survival in LUAD patients
  • PTBP1 expression reduced in five replicative senescent cell lines (BJ, IMR90, WI38, HFF, MRC5)
  • PTBP1 knockdown increased SA-β-Gal staining and induced G2/M arrest with reduced G1 phase
  • PTBP1 knockdown reduced proliferation and migration of A549/H1299 cells
  • PTBP1 depletion decreased FBXO5 mRNA and protein and increased FBXO5-S isoform via exon 3 skipping
  • FBXO5 knockdown induced senescence and G2/M arrest; FBXO5 decreased in five senescence models
  • FBXO5-S degraded significantly faster than FBXO5-L in A549 and H1299
Key statistics
  • count 542 LUAD samples and 59 normal samples (GEPIA dataset for PTBP1 expression/prognosis in LUAD)
  • count 2276 differentially expressed mRNAs (1031 up, 1245 down) (PTBP1-high vs PTBP1-low TCGA-LUAD patients)
  • count 756 alternative splicing events (PTBP1-KD A549 vs control by rMATS)
  • count 355 universally underexpressed genes (across replicative senescence lines MRC5, IMR90, WI38)
  • count 33 cancer types (TIMER analysis of PTBP1 expression)
  • other |log fold change| ≥ 0.5 and adjusted p-value < 0.05 (DEG cutoff threshold (DESeq2))
  • pvalue p < 0.05 (*), p < 0.01 (**), p < 0.001 (***) (significance thresholds, n = 3 replicates, unpaired t-test)

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 combined bioinformatic analyses of public RNA-seq datasets (TCGA-LUAD via GEPIA, GEO GSE63577) with in vitro experiments in A549 and H1299 LUAD cell lines. Differential gene expression in TCGA data was identified with DESeq2, alternative splicing changes were detected by rMATS, and all in vitro group comparisons were performed with unpaired t-tests (n = 3 per group). Results were reported as mean ± SD with symbolic p-value thresholds (*, **, ***).

Replicationbiological Sample sizen = 3 stated for all in vitro experiments; 542 LUAD and 59 normal samples from TCGA/GEPIA for bioinformatic expression analyses; 5 replicative senescence cell lines from GSE63577 GroupsPTBP1-KD (two shRNAs) vs. empty-vector control in A549 and H1299; FBXO5-KD vs. control; PTBP1-high vs. PTBP1-low LUAD patients; senescent vs. proliferating cells; FBXO5-L vs. FBXO5-S isoform mRNA decay over time Pairingunpaired Randomization/blindingnot stated DispersionSD Exact p-valuesno Effect sizesno Confidence intervalsno Multiplicity correctionBenjamini-Hochberg FDR (via DESeq2 adjusted p-value) for bioinformatic DEG analysis; no correction stated for the set of in vitro unpaired t-tests
Statistical tests used
Test Applied to n Assumptions
Unpaired two-sample t-test All in vitro experimental comparisons: proliferation (CCK-8 growth curves), colony formation counts, transwell migration counts, SA-β-Gal quantification, RT-qPCR gene expression levels, mRNA stability (ActD assay time points), and Western blot densitometry n = 3 biological replicates per group (stated in Section 2.11) not stated
DESeq2 Wald test with Benjamini-Hochberg adjusted p-value Differential gene expression between PTBP1-high and PTBP1-low TCGA-LUAD patients stratified by median PTBP1 expression; cut-off |log FC| ≥ 0.5 and adjusted p < 0.05 542 LUAD samples from TCGA retrieved via GEPIA not stated
rMATS (replicate Multivariate Analysis of Transcript Splicing) algorithm Alternative splicing event detection (SE, MXE, A3SS, A5SS, IR) in PTBP1-KD A549 cells vs. control; 756 events identified null not stated
Hypergeometric/Fisher's exact enrichment test (DAVID) KEGG pathway and GO term enrichment for PTBP1-upregulated DEGs in TCGA-LUAD and for universally underexpressed genes across three replicative senescence fibroblast lines (MRC5, IMR90, WI38) null not stated
Survival analysis method not specified (performed via GEPIA web tool) Association of PTBP1 expression with survival rates in LUAD patients 542 LUAD samples from GEPIA/TCGA not stated
Approaches that could also have been used
  • Multiple independent unpaired t-tests were applied across distinct in vitro endpoints (proliferation, migration, SA-β-Gal, cell cycle, RT-qPCR, mRNA stability) within the same study
    Could also: A one-way ANOVA with a post-hoc correction (e.g., Dunnett's test comparing each knockdown to the shared control, or Tukey HSD) could also be applied across the set of comparisons within each experiment — Applying a family-wise or false-discovery-rate correction to a collection of related comparisons is an alternative framework that explicitly accounts for the increased probability of a Type I error when many tests are performed; individual t-tests and ANOVA with post-hoc correction can both be appropriate, and the choice depends on whether the comparisons are treated as a family
  • In vitro results were reported as mean ± SD with n = 3 per group
    Could also: A 95% confidence interval around the mean difference could also be reported alongside or instead of SD — With n = 3, SD describes the sample spread, whereas a CI directly conveys the uncertainty in the estimated group mean; CIs are also directly interpretable for effect magnitude and are increasingly preferred by reporting guidelines for small sample sizes
  • P-values were reported as symbolic threshold categories (* < 0.05; ** < 0.01; *** < 0.001) rather than exact values
    Could also: Exact p-values (e.g., p = 0.018) could also be reported for each comparison — Exact p-values convey the distance from the threshold, support downstream meta-analyses, and allow readers to apply their own significance criteria; threshold symbols alone do not distinguish a p of 0.049 from a p of 0.001
  • PTBP1 expression was dichotomized at the median to define PTBP1-high and PTBP1-low patient groups for DEG and survival analyses
    Could also: Continuous PTBP1 expression could also be used as a covariate in a Cox proportional hazards model for survival, or as a continuous predictor in a linear model for gene expression associations — Median dichotomization converts a continuous variable to binary, which discards expression gradient information and fixes the split at an arbitrary threshold; modeling PTBP1 as continuous avoids threshold dependence and generally retains statistical power
  • The mRNA stability (ActD) assay measured FBXO5-L and FBXO5-S levels at five time points (0, 2, 4, 6, 8 h) and comparisons appear to be made by t-test at individual time points
    Could also: A linear mixed-effects model or repeated-measures ANOVA fitted to the full decay time course could also model the data, with the decay rate (slope) as the primary estimand — Measurements from the same experimental replicate at successive time points are correlated; a longitudinal model accounts for this structure, estimates the half-life directly with a confidence interval, and tests whether the decay curves differ as trajectories rather than at isolated time points
  • DESeq2 was used as the sole method to identify differentially expressed genes in the TCGA-LUAD bioinformatic analysis
    Could also: edgeR (negative-binomial GLM) or limma-voom (precision-weighted linear model) are both well-benchmarked alternatives for bulk RNA-seq differential expression that could also have been applied — Different tools make different distributional assumptions and can yield somewhat different gene lists, particularly at moderate expression levels; reporting the overlap across two independent methods is a common approach to assess the robustness of the identified DEG set
Software: GraphPad Prism 8.3.0 · DESeq2 (R/Bioconductor) 1.26.0 · STAR · StringTie · rMATS · ModFit 3.0 · ImageJ 2.3.0 · DAVID

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
8
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.

What was reproduced

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

Scope — pmid-39057099

Paper: Li et al. 2024, Curr Issues Mol Biol 46(7):458. "Downregulation of Splicing Factor PTBP1 Curtails FBXO5 Expression to Promote Cellular Senescence in Lung Adenocarcinoma." DOI 10.3390/cimb46070458 · PMCID PMC11276454.

Harvested artifacts: code=github.com/alexdobin/STAR (the STAR aligner — a third-party tool, NOT the authors' own repo; per BRIEF rule 2 this is fully valid), data=GEO:GSE63577.

Computational (pipeline-derived) results in the paper

# Result (reported) Pipeline Input data Public? In scope?
C1 355 universally underexpressed genes across replicative-senescence fibroblasts (paper text: MRC5, IMR90, WI38) cross-cell-line DE + intersection of GSE63577 GSE63577 (handed dataset) YES — processed count/RPKM matrices shipped as GEO suppl. .xls.gz PRIMARY
C2 756 alternative splicing events in PTBP1-KD vs control A549 STAR(GRCh37)+StringTie+rMATS, FDR≤0.01, |IncLevelDiff|≥0.1 authors' own A549 PTBP1-KD RNA-seq NO accession given in paper (Data-availability lists only "GSE datasets"); raw fastq not located OUT (data not deposited → would be no_data_accession/data_unavailable for this sub-result)
C3 2276 DEGs (1031 up / 1245 down) PTBP1-high vs PTBP1-low LUAD DESeq2 |logFC|≥0.5, padj<0.05 TCGA-LUAD YES (TCGA public) but PTBP1-high/low grouping threshold not specified SECONDARY (attempt if budget allows; grouping ambiguous)
C4 PTBP1 expression across 33 cancers; LUAD 542 tumor vs 59 normal TIMER / GEPIA web portals TCGA/GTEx via portal partial OUT (portal point-and-click, not a runnable pipeline; non_pipeline)
qRT-PCR, Western blot, SA-β-gal, EdU, growth curves, IHC wet-lab OUT (wet-lab, not computational)

Decision

  • Primary target = C1 (the handed dataset GSE63577). Processed counts are shipped, so the 80% — reproducing the senescence-downregulated signature — is achievable without running STAR on 48 raw libraries (that would be the hard 20%). We reproduce the cross-cell-line "consistently down in senescent vs proliferating" signature and compare its size to the reported 355.
  • Documentation gap (audit-relevant): the paper specifies no threshold, statistical test, or definition of "universally" for the 355-gene set, and is internally inconsistent about whether it spans 3 (MRC5/IMR90/WI38) or 5 (+BJ/HFF) cell lines. The exact integer 355 is therefore not strictly derivable from the shipped data + described methods. We reproduce the methodology under reasonable, explicitly-stated criteria and report how close the count lands — flagging 355 as a possible non-derivable value if it is far off under all reasonable thresholds.
  • C2 dropped for this RU: the A549 PTBP1-KD RNA-seq underlying the 756 splicing events has no GEO/SRA accession in the paper, so it cannot be reproduced from public data. STAR (the "code" link) is exercised conceptually but its input is unavailable.

Compute plan

All on «our HPC» (std partition), data on «infra» «path». Light analysis (download processed matrix + R/Python DE intersection); no --mem flag («infra» rule).

Figures / tables: Fig 4DFig 4Fig 5
C1
Reported
355 universally underexpressed genes across replicative-senescence fibroblasts (Fig 4D)
Reproduced
Not derivable to 355: strict per-cell-line DESeq2 intersection (padj<0.05)=0; combined ~cellline+cond model=4937 (any lfc)/954 (|lfc|>=1); fold-change-only RPKM intersect=690 (3-line>=2x) / 636 (5-line>=1.5x) / 238 (5-line>=2x). Order of magnitude (hundreds) consistent; exact integer not obtainable under any standard criterion.
did not match
C1b
Reported
FBXO5 downregulated in replicative senescence (paper's mechanistic premise)
Reproduced
FBXO5 down in all 5 cell lines, log2FC(old/young) -1.16..-3.14, significant in 4/5 (IMR90 ns)
within tolerance
PC1
Reported
LMNB1 down in senescence (positive control)
Reproduced
LMNB1 log2FC -1.35..-4.45 across all 5 lines, all padj<<1e-30
exact
C2
Reported
756 alternative splicing events in PTBP1-KD A549 (STAR+StringTie+rMATS)
Reproduced
not attempted
partial
C3
Reported
2276 TCGA-LUAD DEGs (1031 up / 1245 down), PTBP1-high vs -low
Reproduced
not attempted
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 59/100

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.

Main result did not reproduce
Decisive
From: Q5 · Derivability / plausibility 🔴

The authors' own deposited dataset GSE63577 reproduces the paper's biology 1:1 qualitatively — FBXO5 and PTBP1 are robustly downregulated in senescence across all 5 fibroblast lines (significant 4/5) and the LMNB1 positive control is strongly down (padj<<1e-30). However, the headline integer '355 universally underexpressed genes' (Fig 4D) is not derivable from the same public matrix under any standard criterion (strict intersection=0; method-dependent 238–4937), because the paper specifies no threshold, no test, no definition of 'universally' and is internally contradictory about whether 3 or 5 lines are intended. This is an authors'-side underspecification (q4/q5 red), not a data-restriction artifact and not proof of fabrication — the order of magnitude (hundreds) holds. Two further claims (756 splicing events, 2276 TCGA DEGs) were untestable due to undeposited data/scope, so the central conclusion is partially confirmed.

🤝
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.

155.4 k
tokens (I/O) · 8.8 M incl. cache
20 min
runtime · 0.05 CPU-h
3.1 GB
peak RAM
4
HPC jobs
hummel
machine