Geng Report: Elevated intracellular copper induces CTR1 monomerization and prevents copper uptake (Nature Communications, 2025; DOI 10.1038/s41467-025-66283-w)

Academic fraud report · Geng Detector

Summary

This Geng-style report raises concerns about a 2025 Nature Communications paper by Wen et al. on CTR1 oligomerization dynamics measured by single-molecule localization microscopy (SMLM). The overall verdict is 'questionable' (yellow). Key issues include: (1) implausibly low variability in SMLM data (SE = 0.04 with n=6 cells, implying SD ≈ 0.098) for weighted mean subunit number ΔN, which is unusually 'textbook-perfect' given cell-to-cell heterogeneity; (2) OCR-style garbled text in Figure 5a and in Gaussian/locational error equations in Methods, raising suspicion of image compositing or pasted formulas; (3) a clear catalog-number typo in Methods where both Alexa Fluor 555- and 647-conjugated WGA are cited as Invitrogen W32466, whereas W32466 corresponds to the 647 conjugate and the 555 conjugate is W32464; (4) a statistical concern that n=6 biological replicates underpin claims based on 1,402,934 localization events. Confidence is limited because no pixel-level image forensics were performed; concerns are inferred from text patterns and reported numbers.

Verdict

� Questionable. Multiple textual and statistical anomalies were identified, but the report is constrained by a lack of pixel-level image analysis. Findings should be treated as leads for formal investigation rather than confirmed misconduct.

Key findings

Implausibly tight SMLM variance (Figure 3d, n=6): WT CTR1 ΔN shifts from 2.30 ± 0.18 to 1.87 ± 0.04 (SE). With n = 6 cells, SE = 0.04 implies SD ≈ 0.098, an unusually narrow distribution for SMLM data subject to cell heterogeneity, expression-level variation, and microenvironmental noise. Severity: 🟠.
OCR-style garbled text in Figure 5a and Methods equations: Western blot marker labels appear as the out-of-order string 37 25 20 15 50 100 150 75 250, and Gaussian / localization-error formulas (Eq. 1 and Eq. 2) display stacked Greek-letter sequences such as e² 1 2 x²x 0 σx Δμ 2 + y²y 0 σy αβ 2 φψφ. Such artifacts are consistent with figure compositing, layer overlays, or pasted equations. Severity: 🟠.
Reagent catalog-number typo (Methods, WGA labeling): Both Alexa Fluor 555- and 647-conjugated WGA are listed as Invitrogen W32466; per Thermo Fisher catalogs, W32466 corresponds to the Alexa Fluor 647 conjugate, while W32464 is the Alexa Fluor 555 conjugate. The duplicated identifier suggests copy-paste error during drafting. Severity: 🟡.
Biological replicate imbalance (Figure 3): Group sizes are n = 6 (WT-Basal), 6 (WT-Cu), 6 (M150L-Basal), 9 (M150L-Cu), while the study analyzes 1,402,934 CTR1 localization events. Statistical power is driven by the n = 6 biological replicates, limiting confidence in reported p-values (e.g., p = 0.0244). Severity: 🟡.
Possible image reuse / region-selection concern (Figure 3 vs Figure 5): WT-Basal trimer fractions of 62.9% (Figure 3c) vs 34.8% in Rab5− regions (Figure 5c) are interpretable as region differences, but combined with Finding 1 they raise the possibility of selective reporting or parameter tuning. Severity: 🟡 (pending raw-image verification).

Evidence highlights

DOI: 10.1038/s41467-025-66283-w.
Reported ΔN values (Figure 3d): WT-Basal 2.30 ± 0.18 → WT-Cu 1.87 ± 0.04, with n = 6 cells per condition.
Total localization events analyzed: 1,402,934 CTR1 molecules.
Anomalous marker sequence observed: 37 25 20 15 50 100 150 75 250.
WGA reagent identifiers cited: Invitrogen W32466 listed for both Alexa Fluor 555 and 647 conjugates.

Notes

All image-related inferences are text-based; no ELA or pixel-level forensics were performed. Conclusive determinations require the original SMLM localization files, uncropped Western blots, and high-resolution figure files.
The n = 6 variance issue is statistically noteworthy but not, on its own, proof of fabrication.
Authors should be contacted for raw data before any formal escalation.