Integrity Review: Mirror-image T7 transcription of chirally inverted ribosomal and functional RNAs

Academic fraud report · Geng Detector

Summary

Verdict: No evidence of academic misconduct detected. This Science 2022 paper by Yuan Xu and Ting F. Zhu (DOI: 10.1126/science.abm0646) reports chemical synthesis of a mirror-image T7 RNA polymerase and transcription of L-RNAs, including a functional mirror-image ribosome. Four dimensions were examined: (1) statistical reporting — p-values span a realistic distribution (P = 0.04, 0.09, 0.1, 0.3, 0.4, 0.7, 0.9, P < 0.0001) rather than clustering suspiciously near 0.05, indicating no p-hacking; (2) methodological logic — authors transparently acknowledge the absence of mirror-image reverse transcriptase/S sequencing and substitute natural-chirality fidelity as a proxy, which is a defensible workaround; (3) data plausibility — although Figure 5 degradation curves for D- vs L-rRNA align almost perfectly with theory, authors honestly note an unpredicted anomaly (band broadening in L-RNA before full degradation) that is hard to fabricate; (4) image analysis — limited by text-only review, but contrast adjustments are disclosed in figure legends. Limitations: pixel-level image forensics and supplementary figures (S1–S73) were not assessed. Overall confidence in this clean verdict is moderate pending raw image inspection.

Verdict

No evidence of academic misconduct. The paper appears authentic based on text-level analysis across statistical, methodological, data-plausibility, and image-disclosure dimensions.

Key findings

Statistical reporting (Figure 4C, Figure 6E): Reported p-values form a realistic spread (0.04, 0.09, 0.1, 0.3, 0.4, 0.7, 0.9, <0.0001). Absence of a suspiciously tight cluster just below 0.05 argues against p-hacking.
Methodological honesty: Authors explicitly state they cannot directly assess mirror-image T7 fidelity due to lack of mirror-image RT and Sanger reagents, and instead use natural-chirality T7 fidelity as a surrogate — a transparent and logically sound substitution.
Timeline consistency: All reagents and references cited (Fmoc-SPPS, NCL, etc.) align with publication date (28 October 2022); no anachronistic or future-only resources detected.
Data plausibility: Figure 5 (A–F) shows strong but expected contrast — D-rRNA fully degrades by 4 h, L-rRNA persists to 720 h (DEPC water) / 168 h (pond water). Crucially, authors document an unplanned observation (peak broadening in L-RNA prior to full degradation), inconsistent with fabricated data.
Image handling: Contrast adjustments are openly disclosed (e.g., Figure 6D: "Contrast-adjusted images are shown on the right..."). No indication of concealment.

Evidence highlights

DOI: 10.1126/science.abm0646
Journal: Science; Year: 2022; Authors: Yuan Xu, Ting F. Zhu
Quantitative anchors preserved: p-values listed above; degradation times 1 h, 4 h, 168 h, 720 h
Figure-level notes: Figure 5 band-broadening anomaly; Figure 6D contrast-adjustment disclosure

Notes

Pixel-level image forensics (gel/Western blot splicing, cloning artifacts) could not be performed from the supplied text-only report.
Supplementary Materials (Figs. S1–S73) and uncompressed original images were not reviewed; a deeper audit would require these files.
Verdict is based solely on textual and statistical patterns; final institutional adjudication would require access to raw data and images.
Disclaimer: this review is AI-assisted and intended for academic discussion only.