Verdict
Highly suspicious. The principal concerns are severe methodological inconsistencies and logical flaws in experimental design. No direct evidence of data fabrication was detected, but the described practices are insufficient to support the quantitative claims made.
Key findings
- Use of G0 mosaics for all downstream analyses: Section 3.3 explicitly states "This prompted us to use G0 mutants for subsequent analyses." Using G0 mosaic adults for quantitative phenotyping and RNA-seq in Lepidoptera is a recognized methodological flaw, because somatic and germline tissues carry heterogeneous indel mosaics. The reported PCA figures (PC1 = 56.95%, PC2 = 13.52%) are consistent with the high inter-individual variance expected from pooled mosaics.
- Implausible 100% mutation rate in G0: The claim that 20 randomly selected G0 adults all carried the expected mutation is atypical for NHEJ-dominated insect CRISPR editing and warrants raw Sanger chromatograms for verification.
- Internal methodological contradiction: Section 2.6 states sperm were released into 100 µL PBS, then in the next sentence into 400 µL PBS. This inconsistency directly undermines the reliability of the reported 64.3% total sperm reduction.
- Repetitive publication pattern: Two cited references (Ref 29, Wang P. 2025, Pest Manag Sci; Ref 36, Zhao Y. 2025, Insect Biochem Mol Biol) follow the same narrative template (testis-specific gene → CRISPR knockout → ~20% hatch rate → ~64% sperm reduction → transcriptomic pathway enrichment). This does not prove misconduct but indicates a rigid, potentially automated production pattern.
- DOI: 10.1002/ps.70871
- Direct quote (Section 3.3): "This prompted us to use G0 mutants for subsequent analyses."
- PCA values: PC1 = 56.95%, PC2 = 13.52%
- Section 2.6 contradiction: 100 µL PBS vs. 400 µL PBS in adjacent sentences
- Reported phenotypic effect: 64.3% reduction in total sperm count
- Hatch rate claim: ~20% in mutant crosses
- Related same-group papers: Wang P. 2025 (Pest Manag Sci, Ref 29); Zhao Y. 2025 (Insect Biochem Mol Biol, Ref 36)
- Image-level analysis (Figure 1–Figure 6) was not possible because only annotated PDF text was available; no pixel-level checks for splicing, noise patterns, or duplicated panels could be performed.
- The chimeric nature of G0 individuals means individual-level variation in mutation spectra is expected; pooling these for RNA-seq is statistically problematic.
- Confidence in the methodological flaws is high (text-based). Confidence in any intent to fabricate is low; genuine experimental sloppiness or overselling of preliminary data is an equally plausible explanation.
- Recommended actions: request original G0 Sanger traces, request G1/G2 homozygote validation data, and raise the Section 2.6 volume contradiction on PubPeer for clarification.