See the raw experimental evidence behind an author's publications and reproducibility signals.
Press Enter ↵ to solve
Fuel Your Discoveries
"The greatest enemy of knowledge is not ignorance, it is the illusion of knowledge."
- Stephen Hawking
Quick Explanation
Copied
Kailiang Qiao — evidence-based science audit
Across the provided publication list, Qiao’s work is heavily centered on hepatocellular carcinoma biology (EMT, angiogenesis/VM, m6A/Hippo, DUB/SLUG), with additional mechanistic cancer–immunity links (e.g., exosome/cytokine storm) and an explicitly retracted VM-related paper in the record. This suggests both scientific productivity and the need for careful scrutiny of mechanistic claims and reproducibility, especially around VM assays and drug-synergy interpretations.
Scope of evidence used: Only the publication metadata/DOIs and study-type summaries explicitly provided in the user message, plus the additional raw-data entry describing an ecDNA–cGAS–STING paper. No unprovided claims about the author’s CV or lab output are made.
1) Research themes (from provided papers)
EMT regulation in HCC via USP5 stabilizing SLUG: mechanistic axis from DUB activity → transcriptional EMT driver → functional phenotypes. (10.7150/thno.27654)
Vasculogenic mimicry (VM) / angiogenic plasticity in HCC via m6A–Hippo signaling. (10.1007/s10456-020-09744-8)
Extracellular vesicle/exosome immunomodulation for cytokine storm reduction in sepsis (biomimetic immunosuppressive exosomes). (10.1002/adma.202108476)
HCC microenvironment & drug resistance linking hypoxia/HIF-1α to platinum resistance using salidroside as a modulator. (10.1016/j.ebiom.2018.10.069)
Drug-adjunct/EMT modulation via dihydroartemisinin (DHA) and Akt–Snail signaling. (10.18632/oncotarget.21793)
Important quality signal: The provided list includes a retracted VM-targeting study: 10.1038/s41419-018-0902-5 (explicitly marked as retracted in the provided list). Retractions are not automatically “author misconduct,” but they materially increase uncertainty around at least some experimental claims in that line.
2) Visual evidence: ecDNA–cGAS–STING barrier paper (raw-data entry provided)
The user-provided raw-data entry describes an ecDNA–innate immunity study with explicit sample sizes and methods. Below, I visualize the tumor group sizes mentioned and the experimental strategy breadth (in vitro + in vivo).
(This section is about the paper described in the raw-data entry, not a claim that Qiao authored that specific work.)
Critical appraisal of the ecDNA–cGAS–STING claims (from provided raw-data entry)
Strengths (as stated): multi-level evidence across tumor cohorts, human cell lines, and mouse models; mechanistic assays including cGAMP quantification (LC-MS), cGAS localization/condensates, and interventions (decitabine, cGAMP delivery, cGAS expression). (10.64898/2025.12.31.697191)
Uncertainties / potential blind spots (as stated): ecDNA abundance in vivo is low and imaging is difficult; downstream nodes beyond cGAS/STING may regulate outcomes; precise mechanism of how STING suppresses de novo ecDNA biogenesis is not fully resolved; translatability across ecDNA+ heterogeneity requires further validation. (10.64898/2025.12.31.697191)
I cannot verify experimental details beyond what is implied by titles/metadata in the provided list. So this section is a skeptical structured audit focusing on: (i) mechanistic specificity implied by the title, (ii) whether the record includes corrections/retractions, and (iii) whether the work spans orthogonal modalities (cell biology, signaling, functional assays)—noting that the latter must be confirmed from full text.
Skeptical synthesis across the provided titles
Mechanistic specificity is repeatedly claimed via signaling-node naming (e.g., HIF-1α, Hippo, Akt–Snail), which is generally a positive sign for falsifiability if the experiments include causal perturbations and orthogonal readouts. However, without full-text methods/results here, evidence strength cannot be upgraded beyond “plausibility from mechanistic framing.” (e.g., 10.1007/s10456-020-09744-8, 10.1016/j.ebiom.2018.10.069, 10.18632/oncotarget.21793)
VM research is high-risk for measurement/interpretation artifacts (assay definition, imaging criteria, and whether “VM” corresponds consistently to a reproducible biological entity). The presence of an explicitly retracted VM-targeting paper in the record increases the need for methodological scrutiny for the VM-related mechanistic chain. (10.1038/s41419-018-0902-5)
Corrections/corrigenda exist in the list (a corrigendum is shown in the provided metadata list for a related EBioMedicine paper). Corrections are not inherently negative, but they do require readers to check whether they affect key mechanistic conclusions or only minor errors. (Corrigendum metadata referenced in provided list; specific DOI not provided in your message for that corrigendum.)
Breadth across modalities appears plausible (signaling, EMT, angiogenesis, exosome immunology). But methodological breadth (e.g., proteomics, in vivo kinetics, pre-registered analysis) cannot be confirmed without full text. Therefore, I rate the current audit as title/metadata-based, not full experimental verification.
4) What would most improve this author review (disproof targets)
To strengthen or refute the author’s mechanistic claims, the most decisive missing elements are full-text experimental details for each key claim-line. Specifically, I would want:
For EMT/SLUG (USP5): causality strength—e.g., rescue experiments, multiple independent perturbations of USP5, and specificity controls to exclude off-target DUB effects. (10.7150/thno.27654)
For VM (m6A–Hippo and VM-inhibitory claims): assay operationalization (criteria for VM), blinded quantification, and whether changes generalize across cell lines and conditions. (10.1007/s10456-020-09744-8, 10.1038/s41419-018-0902-5)
For exosome sepsis claims: biodistribution, dose–response, and whether cytokine changes reflect direct immunomodulation vs confounded viability/clearance effects. (10.1002/adma.202108476)
For hypoxia/platinum resistance (HIF-1α) and Akt–Snail EMT suppression: whether pathway perturbations are consistent with a single causal mechanism (and not only correlation with drug response metrics). (10.1016/j.ebiom.2018.10.069, 10.18632/oncotarget.21793)
Confidence level of this audit: moderate for “topic mapping from titles/metadata,” low for “experimental conclusion strength,” because full methods/results and the specific retraction/correction scope are not included in the provided text.
Explore deeper with BGPT (targeted actions)
Run an AI Scientist analysis (iterative, computation-enabled)
This will iteratively check the provided publication DOIs, extract experimental evidence structure (where available), and produce a stricter uncertainty map across EMT/VM/drug-resistance/exosome claims.
Feedback:
Updated: May 01, 2026
BGPT Author Review
Scientific Quality
60%
Based on the provided publication titles/metadata alone, the author appears to focus on mechanistic cancer signaling (EMT, Hippo/m6A, Hsp90β angiogenesis, HIF-1α drug resistance) and includes at least one explicitly retracted VM-related paper in the record, which increases uncertainty about that research line. However, the audit cannot verify experimental rigor, causality, blinding, replication, dataset/controls, or whether corrections/retractions affect core conclusions because full-text evidence is not included in the prompt. Overall: promising mechanistic framing with a notable reliability risk signal (retraction) and insufficient accessible methodological detail for a higher confidence score.
Communication Quality
60%
From titles/metadata, the work communicates specific pathway/node hypotheses (e.g., HIF-1α, Akt–Snail, Hippo) which is good for clarity. But actual writing quality, figure clarity, and whether limitations are transparently discussed cannot be evaluated without abstracts/full text in the prompt, limiting confidence.
Author Novelty
50%
The research themes (EMT regulation, VM, crosstalk via Hippo/m6A, HIF-1α hypoxia drug resistance, exosome immunomodulation) are established topic areas. The novelty likely depends on the specific causal links and experimental design, which are not verifiable here beyond mechanistic naming. Therefore novelty is judged as moderate with uncertainty.
Scientific Rigor
50%
Mechanistic specificity implied by titles suggests an attempt at causal biology rather than purely correlative work. Still, rigor cannot be confirmed (e.g., number of independent experiments, rescue controls, assay standardization for VM, blinding/quantification). The presence of a retracted VM paper increases the chance that at least some VM-related measurements/interpretations may have been insufficiently reliable, lowering the rigor estimate from title-level evidence alone.
Quick Explanation
Long Explanation
Top Data Sources
ExportMCP
Discussion
Fuel Your Discoveries
Ask a Follow-Up
Key Insight