Integrity Assessment: "RBMS1 Coordinates with the m6A Reader YTHDF1 to Promote NSCLC Metastasis through Stimulating S100P Translation" (Adv. Sci., 2024)

Academic fraud report · Geng Detector

Summary

Verdict: Cleared based on available evidence (✅ / 清白). This textual audit of the paper by Sun et al., DOI 10.1002/advs.202307122, found no indicators of fraud across the dimensions that could be evaluated. Mathematical verification of Table 1 (n=40) reconciled exactly across age, gender, and lymph-node metastasis strata; the reported 68.8% for positive/high lymph-node metastasis equals 11/16 = 68.75%. Ethics approval dates (animal: AEE21015 ~2021; human: PJ-KS-KY-2022-208 ~2022) precede the submission date (September 26, 2023), with no fabrication signals. Reagent choices (Nortriptyline repurposed as an RBMS1 inhibitor, justified by the authors' prior J Clin Invest 2021 work) and culture media (F12K for A549, RPMI-1640 for H460/H2170, DMEM for HEK293T) match ATCC recommendations. Statistical tests declared in the methods (one-way ANOVA with Tukey/Dunnett, t-tests for pairwise) align with the figure legends. Limits: pixel-level image-duplication and splicing analyses could not be performed because only the PDF text was available; Figures 3 (Co-IP) and 4 (Western Blots) remain unverified at the image level.

Verdict

Cleared based on textual, logical, methodological, and timeline analysis. No fraud indicators detected in the evaluated dimensions.

Key findings

Table 1 arithmetic verified: 40-patient cohort; age strata (≤65: 17; >65: 23), gender (Male 32, Female 8), and lymph-node metastasis (Negative 24, Positive 16) sum correctly. Positive/High proportion = 11/16 = 68.75%, matching the reported 68.8%.
P-value distribution is plausible: Clinical stage P=0.0491 and lymph-node metastasis P=0.028 fall within a reasonable spread; no clustering at the α=0.05 threshold suggestive of p-hacking.
Ethics timeline is consistent: Animal ethics ID AEE21015 (~2021) and human sample ethics ID PJ-KS-KY-2022-208 (~2022) both precede the manuscript submission (Received: September 26, 2023). The numbering format is consistent with Dalian Medical University conventions; no fabrication signs observed.
Reagents are legitimate: Nortriptyline hydrochloride (NTP) is an established drug repurposed as an RBMS1 inhibitor, supported by the authors' own prior work (J Clin Invest 2021, Ref 15).
Cell culture media match ATCC recommendations: A549 in F12K, H460/H2170 in RPMI-1640, HEK293T in DMEM — no errors such as misclassifying suspension cells as adherent.
Statistical methods are internally consistent: Methods section specifies one-way ANOVA with Tukey/Dunnett for multi-group and t-tests for pairwise comparisons; figure legends for Figs. 1–3 and 6 use the same tests.
Differential cell seeding reported: A549 at 5×10⁴ vs. H460 at 1×10⁵ in Transwell assays — a level of detail consistent with genuine empirical optimization rather than copy-paste fabrication.

Evidence highlights

Table 1 (Page 2): 40-sample reconciliation across all three demographic/clinical variables; lymph-node positive/high = 11/16 = 68.75%.
Ethics IDs (Page 14): AEE21015 (animal), PJ-KS-KY-2022-208 (human).
Submission timeline: Received September 26, 2023 — both ethics approvals precede this date.
Reagent citation (Ref 15): J Clin Invest 2021, supporting NTP as RBMS1 inhibitor.
Methods vs. figure legends: One-way ANOVA + Tukey/Dunnett; t-test for pairwise comparisons — matched across methods section and Fig. 1/2/3/6 legends.

Notes

Scope limitation: This audit was performed on extracted PDF text only. Pixel-level image-duplication (Figures 1–7), background-texture analysis, and splicing checks for Co-IP (Figure 3) and Western blots (Figure 4) were not possible. A definitive clearance would benefit from ImageJ/Forensically-based reanalysis of the original high-resolution figures.
Cross-paper image reuse: The authors' prior publication record (including Ref 15, J Clin Invest 2021) was not screened for inter-paper figure reuse; this is a common target in image-fraud investigations.
Confidence: High confidence for textual/logical integrity; low-to-moderate confidence for overall paper integrity pending image-level verification.
Disclaimer: This report is AI-assisted and intended for academic discussion only. Final determinations of misconduct require formal institutional investigation.