Investigation Report: Suspected Data Manipulation and Methodological Irregularities in 'Inhibition of Cell Proliferation and Migration through Nucleobase-Modified Polyamidoamine-Mediated p53 Delivery' (DOI: 10.2147/IJN.S146917)

Academic fraud report · Geng Detector

Summary

This report evaluates a 2018 International Journal of Nanomedicine paper by Haobo Han et al. on AP-PAMAM-mediated p53 delivery for cervical cancer, and renders a verdict of 'highly suspicious' on multiple grounds. The investigation identified five principal concerns: (1) a highly modular publication pattern, where the same group's 2017 Biomater Sci paper (DOI-linked, ref. 31) used the identical AP-PAMAM delivery system for miR-23b in lung cancer, indicating likely replication of materials, methods, and experimental design; (2) a physically implausible V-shaped discontinuity in the N/P-dependent hydrodynamic diameter data in Table S1 (125.6 nm at N/P=15; 143.2 nm at 25; 108.9 nm at 30; 141.8 nm at 40; 199.7 nm at 50), with the N/P=30 minimum coinciding suspiciously with the authors' claimed 'optimal ratio'; (3) a residual copy-paste error in Methods ('DEME' instead of DMEM); (4) text parsing limitations preventing full image-level forensics, though Western blot and flow cytometry panels warrant raw-data verification; and (5) an arithmetic inconsistency between Figure 4 quadrant values (13.95% + 8.18% = 22.13%) and the apoptosis rate stated in the main text (26.17%), a discrepancy of nearly 4 percentage points. These findings suggest probable data fabrication or selective reporting rather than genuine independent research. Confidence is high for concerns 1–3 and 5; image-based concerns remain pending raw-data inspection.

Verdict

🟠 Highly suspicious. The paper exhibits multiple convergent red flags consistent with a "publication factory" workflow: near-identical methodology recycled from the same group's 2017 publication, physically implausible physicochemical data trends at the authors' claimed optimum, a quantitative mismatch between figure-derived and text-reported apoptosis rates, and residual copy-paste artifacts in the Methods section.

Key findings

Modular / "salami" publication pattern (Finding 1): The AP-PAMAM synthesis, characterization workflow, and cellular assay architecture mirror the authors' own 2017 Biomater Sci paper (ref. 31) that delivered miR-23b for lung cancer. Only the cargo (p53) and cell line (cervical cancer) appear to differ.
Implausible nanoparticle-sizing data (Finding 2): Hydrodynamic diameter shows an anomalous dip at N/P=30 (108.9 nm) flanked by larger values at N/P=25 (143.2 nm) and N/P=40 (141.8 nm), contradicting the expected monotonic compression-then-plateau behavior of polycation/DNA complexes.
Methods-section typo indicating template reuse (Finding 3): "DEME" appears where DMEM is meant, a tell-tale copy-paste residue.
Image forensics limited (Finding 4): Pixel-level inspection of Figure 3, Figure 6, and Figure S6 was not possible from the parsed text; however, the uniform four-column layout of Western blot bands (Procaspase-3, -8, -9, p53, β-actin) in Figure 6 raises legitimate questions about membrane provenance and potential splicing.
Internal numerical inconsistency (Finding 5): Figure 4 quadrant values (13.95% + 8.18%) sum to 22.13%, whereas the main text reports an apoptosis rate of 26.17%—a ~4 percentage-point gap suggesting either non-standard quadrant gating or post-hoc numerical adjustment.

Evidence highlights

| Item | Value cited in paper | Expected / computed | |---|---|---| | Hydrodynamic diameter at N/P=15 | 125.6 nm | – | | Hydrodynamic diameter at N/P=25 | 143.2 nm | – | | Hydrodynamic diameter at N/P=30 | 108.9 nm | Should not be a localized minimum | | Hydrodynamic diameter at N/P=40 | 141.8 nm | – | | Hydrodynamic diameter at N/P=50 | 199.7 nm | – | | Apoptosis rate (Figure 4 quadrants) | 13.95% + 8.18% = 22.13% | – | | Apoptosis rate (main text) | 26.17% | Discrepancy of ~4.04 percentage points | | Methods text | "DEME" | Should be DMEM | | Related prior publication | Han et al., Biomater Sci. 2017;5(11):2268–2275 (cited as ref. 31) | – |

Notes

DOI: 10.2147/IJN.S146917
Confidence: high for textual/arithmetical findings (1, 2, 3, 5); low-to-moderate for image-based concerns (4) due to text-only extraction.
Recommended follow-up: (a) cross-compare raw Western blot images and FCS files between this paper and Biomater Sci. 2017;5(11):2268–2275; (b) request triplicate DLS measurements underlying the N/P=30 data point; (c) request original FCS files to reconcile the 22.13% vs. 26.17% discrepancy.
This report is AI-assisted and intended for academic discussion; final determination of misconduct requires formal institutional investigation.
No findings have been invented; all numeric values are preserved as they appear in the source report.