Spot-check this review
Roughly half the results here were adjudicated by a single reviewer who was not blinded to the pipeline's provisional labels. Rather than ask you to take that on trust, this page publishes a fixed sample of results — with the source, the page and the sentence for each — so anyone can verify them independently.
What to check
Open the paper, find the quoted sentence, then ask four questions in order and stop at the first failure:
- Is the quote really in this paper, and is it this paper's own result rather than a figure it quotes from someone else?
- Does the statement say what the quote says — no more, no less?
- Do the value, unit and denominator match the quote, and is the rate the right way round?
- Is this genuinely an animal-versus-human comparison, with the right species and the right evidence level?
Not failures: paraphrase that preserves the meaning, rounding, or a page number one off from where you find the sentence. Failures worth reporting loudly: questions 1 and 4 — those mean the result should never have been kept, and they tend to come in classes rather than singly.
Two samples, drawn by fixed seeds so they can be redrawn and audited. Sample A is random across all kept results; sample B is drawn only from the single-reviewer studies, and is the one to do first. Reporting which question failed is more useful than prose: it says whether to re-extract, re-adjudicate a category, or correct one row.
Sample A — random across all kept results (30)
A1. Zebrafish toxicological screening could aid Leishmaniosis drug discovery
2021 · Laboratory Animal Research · PubMed · DOI · study page · page 7 · adjudication: single reviewer
The zebrafish model did not predict the side effects of carbamazepine and benznidazole, which are common in humans.
species zebrafish · level B · Discordant
“Thus, the proposed platform was sensible to perform an effective screening for approval of new compounds, however, it did not predict the side effects of the drugs.”
A2. Patient-derived xenografts effectively capture responses to oncology therapy in a heterogeneous cohort of pati
2017 · Annals of Oncology · PubMed · DOI · study page · page 7 · adjudication: single reviewer
The sensitivity for the PDX drug screens was 96% (80/83).
species mouse · level A · Concordant · 96 percent (n 80/83)
“Our analysis revealed a sensitivity for the PDX drug screens of 96%, with a lower 95% confidence interval (CI) of 89% (which meets the acceptance criteria of 80%), and specificity of 70%, with a lower 95% CI of 54% (Table 1, top panel).”
A3. Plasma-to-tumour tissue integrated proteomics using nano-omics for biomarker discovery in glioblastoma
2025 · Nature Communications · PubMed · DOI · study page · page 1 · adjudication: single reviewer
48 potential glioblastoma biomarker candidates were identified through cross-species correlation.
species grouped-label · level C · Concordant · 48 count
“Cross-species correlation identified 48 potential GB biomarker candidates involved in actin cytoskeleton organisation, focal adhesion, platelet activation, leukocyte migration, amino acid biosynthesis, carbon metabolism, and phagosome pathways.”
A4. Cntnap2-dependent molecular networks in autism spectrum disorder revealed through an integrative multi-omics a
2022 · Molecular Psychiatry · PubMed · DOI · study page · page 3 · adjudication: single reviewer
Out of 7413 common genes between mouse proteomic data and human transcriptome data, 48 genes showed the same expression trend.
species grouped-label · level C · Concordant · 48 count (n 48/7413)
“Of the 7413 genes found to be common to both datasets (Fig. 4a), 48 genes (12 upregulated and 36 downregulated DEGs) showed the same trend as the DEPs in our proteomic data (Fig. 4b).”
A5. Identification of conserved canonical marker genes in human and mouse adrenal glands using Visium spatial tran
2025 · Histochemistry and Cell Biology · PubMed · DOI · study page · page 14 · adjudication: single reviewer
The Spearman correlation of log2FC values between human ZR and mouse ZF across 18 signature genes was 0.4.
species grouped-label · level C · Concordant · 0.4 correlation
“The correlations for other clusters were considerably lower (ρ=0.4 for ZF, ZG, and medulla; ρ=−0.8 for CT/WAT), yielding a specificity margin Δρ=0.4 (Fig. 7c).”
A6. Know Thy Model: Charting Molecular Homology in Stromal Reprogramming Between Canine and Human Mammary Tumors
2019 · Frontiers in Cell and Developmental Biology · PubMed · DOI · study page · page 7 · adjudication: model
The paper found strong molecular homology in stromal reprogramming between canine and human mammary carcinomas.
species grouped-label · level C · Concordant
“In conclusion, these results clearly demonstrated that stromal reprogramming in canine and human mCA shares significant molecular homology.”
A7. Cross-species single-cell landscapes identify the pathogenic gene characteristics of inherited retinal disease
2024 · Frontiers in Genetics · PubMed · DOI · study page · page 12 · adjudication: single reviewer
The IRDs GNAT2, PDEFG, and CSPG4 displayed highly consistent expression in humans and mice.
species grouped-label · level C · Concordant
“The IRDs (GNAT2, PDEFG, and CSPG4) displayed highly consistent expression in humans and mice (Supplementary Figure 5F).”
A8. Concordance of the Toxicity of Pharmaceuticals in Humans and in Animals
2000 · Regulatory Toxicology and Pharmacology · PubMed · DOI · study page · page 5 · adjudication: model
The true positive human toxicity concordance rate for nonrodent species was 63% for human toxicity events.
species grouped-label · level B · Concordant · 63 percent (n /221)
“Concordance was seen in 63% of nonrodent studies (primarily the dog) and 43% of rodent studies (primarily the rat).”
A9. Clonogenic assay with established human tumour xenografts
2004 · European Journal of Cancer · PubMed · DOI · study page · page 1 · adjudication: single reviewer
97% of 80 comparisons between clinical response of tumors and their explants established in nude mice and treated in vivo showed correct prediction for tumor resistance.
species mouse · level A · Concordant · 97 percent (n /80)
“Of 80 comparisons performed, we observed a correct prediction for tumour resistance in 97% and for tumour sensitivity in 90%.”
A10. Tripartite factors leading to molecular divergence between human and murine smooth muscle
2020 · PLoS ONE · PubMed · DOI · study page · page 10 · adjudication: model
19% of EC non-uniformly conserved proteins were absent in mouse.
species mouse · level C · Discordant · 19.0 percent
“Non-uniformly conserved VSMC proteins were much more likely to be absent in mouse than EC or VSMC/EC non-conserved molecules (84.6% for VSMC, compared with 19% for EC proteins and 42.1% for VSMC/EC proteins).”
A11. Patient-derived xenografts effectively capture responses to oncology therapy in a heterogeneous cohort of pati
2017 · Annals of Oncology · PubMed · DOI · study page · page 8 · adjudication: single reviewer
When the first systemic therapy was received after tumor collection, the sensitivity of PDX drug screens was 97% (29/30).
species mouse · level A · Concordant · 97 percent (n 29/30)
“Table 1. Values and confidence intervals for parameters of analytical and clinical accuracy for PDX model drug screening ... Treatment is first therapy received after tumor collection ... Sensitivity 97% (29/30) 81%-99%”
A12. Translating medication effects for alcohol use disorder across preclinical, human laboratory, and clinical tri
2025 · Translational Psychiatry · PubMed · DOI · study page · page 5 · adjudication: single reviewer
Medications that reduced alcohol preference in preclinical 2-BC studies were associated with lower return to drinking in human clinical trials, with a linear slope of 0.04.
species grouped-label · level A · Concordant · 0.04
“For the preclinical 2-BC preference endpoint, the linear slope is statistically significant for return to any drinking in RCTs (β = 0.04, SE = 0.02, p=0.004; Fig. 2B), such that medications that reduced alcohol preference in preclinical 2-BC studies were associated with lower return to drinking in RCTs.”
A13. An integrated analysis of genes and functional pathways for aggression in human and rodent models
2018 · Molecular Psychiatry · PubMed · DOI · study page · page 7 · adjudication: model
The pathway overlap between human GWAS and rodent transcriptome genes was not significant, with 27 pathways in common.
species grouped-label · level C · Discordant · 27 count
“In contrast, the pathway overlap between the human GWAS and rodent transcriptome genes was not significant (N = 27, p = 0.20).”
A14. Expression profiles of the autism-related SHANK proteins in the human brain
2023 · BMC Biology · PubMed · DOI · study page · page 12 · adjudication: single reviewer
SHANK2 puncta density and sum intensity in the amygdala were considerably higher in mice compared to humans.
species grouped-label · level C · Discordant
“In contrast, puncta density and sum intensity in the AMY were considerably higher in mice compared to humans.”
A15. Complement response to burn injury: systematic review and meta-analysis of patient and animal studies
2026 · Frontiers in Immunology · PubMed · DOI · study page · page 8 · adjudication: single reviewer
Human studies showed markedly higher levels of C3a compared with animal studies, with a standardized mean difference of 2.18 in humans.
species non-human-primate · level A · Mixed · 2.18 (n /9)
“Compared with animal studies, human studies showed markedly higher levels of C3a (animals: SMD = -1.34; CI95% = -1.78, -0.89; n = 1 study versus humans: SMD = 2.18; CI95% = 1.79, 2.56; n = 9 studies; P< 0.0001)”
A16. Transcriptomic evidence for immaturity of the prefrontal cortex in patients with schizophrenia
2014 · Molecular Brain · PubMed · DOI · study page · page 8 · adjudication: model
110 genes overlapped between human developing MFC and Shn-2 KO mice MFC, indicating significant similarities in the pattern of transcriptome changes between the two groups (P = 0.0063).
species mouse · level C · Concordant · 0.0063
“One hundred and 10 genes overlapped between human developing MFC and Shn-2 KO mice MFC, indicating significant similarities in the pattern of transcriptome changes between the two groups (P = 0.0063, Figure 4a, Additional file 1: Table S22).”
A17. Amphetamine increases motivation of humans and mice as measured by breakpoint, but does not affect an Electroe
2024 · Cognitive Affective & Behavioral Neuroscience · PubMed · DOI · study page · page 1 · adjudication: single reviewer
Amphetamine increased breakpoint in mice (cohort 2, without EEG tethering) at 0.3 mg/kg.
species mouse · level A · Concordant
“In cohort 2, however, 0.3 mg/kg of amphetamine increased breakpoint consistent with human findings.”
A18. A multimodal cross-species comparison of pancreas development
2025 · Nature Communications · PubMed · DOI · study page · page 1 · adjudication: single reviewer
Transcription factors regulated by NEUROG3 were over 50% conserved between pig and human.
species grouped-label · level C · Concordant · 0.5 percent
“Transcription factors regulated by NEUROG3, the endocrine master regulator, are over 50% conserved between pig and human, many being validated in human stem cell models.”
A19. The dog as an animal model for bladder and urethral urothelial carcinoma: Comparative epidemiology and histolo
2018 · Oncology Letters · PubMed · DOI · study page · page 6 · adjudication: model
Urothelial carcinoma is the most common type of bladder cancer in both people and dogs with comparable frequencies.
species grouped-label · level C · Concordant
“UC is the most common type of bladder cancer in both people and dogs with comparable frequencies.”
A20. Transcriptomic Cross‐Species Analysis of Chronic Liver Disease Reveals Consistent Regulation Between Humans an
2021 · Hepatology Communications · PubMed · DOI · study page · page 15 · adjudication: single reviewer
The highest precision for human NAFLD was 0.33 in the Western-type diet mouse model.
species mouse · level C · Concordant · 0.33
“Among the 12 mouse models analyzed, the WTD presented with the highest precision (0.33) for human NAFLD.”
A21. Bridging the Gap Between Fluid Biomarkers for Alzheimer’s Disease, Model Systems, and Patients
2020 · Frontiers in Aging Neuroscience · PubMed · DOI · study page · page 1 · adjudication: model
The core Alzheimer's disease biomarkers have been found to translate well across species.
species grouped-label · level C · Concordant
“The core AD biomarkers have been found to translate well across species, whereas biomarkers of neuroinflammation translate to a lesser extent.”
A22. Acquired resistance to the RAS(ON) multi-selective inhibitor daraxonrasib guides rational combination therapy
2026 · Nature Medicine · PubMed · DOI · study page · page 5 · adjudication: single reviewer
The majority of key acquired resistance mechanisms identified in human patient ctDNA samples were recapitulated in daraxonrasib-resistant preclinical models.
species grouped-label · level A · Concordant
“The majority of key acquired resistance mechanisms identified in the EOT ctDNA samples (for example, mutant KRASamp, RTK alterations and MYCamp) were recapitulated in daraxonrasib-resistant preclinical models derived following prolonged RAS(ON) inhibition.”
A23. Transcriptomic classification of genetically engineered mouse models of breast cancer identifies human subtype
2013 · Genome biology · PubMed · DOI · study page · page 7 · adjudication: model
The murine NeuEx class associated with the human luminal A subtype based on gene set analysis.
species mouse · level C · Concordant
“Previously defined as a 'luminal' model [31], the Neu Ex murine class associated with the human luminal A sub-type in this newest analysis; this correlation was somewhat surprising given the lack of ERa and ERa-regulated gene expression in the murine NeuEx class, but does suggest that human luminal A tumors have many ERa-independent features.”
A24. Phosphoproteomics reveals conserved exercise‐stimulated signaling and AMPK regulation of store‐operated calciu
2019 · The EMBO Journal · PubMed · DOI · study page · page 4 · adjudication: model
190 phosphosites were regulated in the same region of the protein across human, rat, and mouse datasets.
species grouped-label · level C · Concordant · 190 count
“Mapping potential orthologs by sequence window resulted in the identification of 190 phosphosites regulated in all three models (114 up- and 76 down-regulated phosphosites), representing 80 proteins (Table EV2).”
A25. Plasma Chemokines in Patients with Alcohol Use Disorders: Association of CCL11 (Eotaxin-1) with Psychiatric Co
2017 · Frontiers in Psychiatry · PubMed · DOI · study page · page 13 · adjudication: model
The paper found that changes in plasma CXCL12 concentrations in male rats exposed to repeated ethanol were similar to those observed in humans during abstinence.
species grouped-label · level A · Concordant
“studies conducted in male rats exposed to repeated ethanol revealed similar changes in plasma CXCL12 concentra- tions than those observed in humans during abstinence during abstinence and acute ethanol exposure also induced a decrease in CXCL12”
A26. Comparative HPLC-MSn analysis of canine and human meibomian lipidomes: many similarities, a few differences
2011 · Scientific Reports · PubMed · DOI · study page · page 11 · adjudication: model
Canine meibum had a relatively larger proportion of (O-acyl)-omega-hydroxy fatty acids (OAHFA) than human meibum.
species laboratory-dog · level C · Discordant
“Canine meibum has a relatively larger proportion of OAHFA (minor HPLC with RT 4–6 min, Fig.1A) than human meibum.”
A27. A comparative encyclopedia of DNA elements in the mouse genome
2014 · Nature · PubMed · DOI · study page · page 4 · adjudication: model
79.6% of chromatin-based promoter predictions in the mouse genome have homologues in the human genome with at least 10% overlapping nucleotides.
species mouse · level C · Concordant · 79.6 percent
“This analysis showed that 79.3% of chromatin-based enhancer predictions, 79.6% of chromatin-based promoter predictions, 67.1% of the DHS, and 66.7% of the transcription factor binding sites in the mouse genome have homologues in the human genome with at least 10% overlapping nucleotides, while by random chance one expects 51.2%, 52.3%, 44.3% and 39.3%, respectively (Fig. 3a, Supplementary Information for details).”
A28. Comparative developmental genomics of sex-biased gene expression in early embryogenesis across mammals
2023 · Biology of Sex Differences · PubMed · DOI · study page · page 5 · adjudication: single reviewer
The orthologs of 17 transcription factors expressed in mouse were detected in human embryos.
species grouped-label · level C · Mixed · 17 count
“Surprisingly, the orthologs of only 17 TFs and 6 EEs expressed in the mouse were detected in human embryos.”
A29. Know Thy Model: Charting Molecular Homology in Stromal Reprogramming Between Canine and Human Mammary Tumors
2019 · Frontiers in Cell and Developmental Biology · PubMed · DOI · study page · page 7 · adjudication: model
Upregulated genes in canine normal stroma were highly enriched among favorably prognostic genes in humans.
species grouped-label · level C · Concordant
“And finally, we demonstrated that the high level of molecular homology between canine and human stromal reprogramming manifested in a prognostic value of the canine CAS signature, with upregulated genes in canine CAS highly enriched among adversely prognostic genes in humans, and upregulated genes in canine normal stroma highly enriched among favorably prognostic genes in humans.”
A30. Concordant and discordant gene expression patterns in mouse strains identify best-fit animal model for human t
2017 · Scientific Reports · PubMed · DOI · study page · page 7 · adjudication: model
There were no enriched discordant modules in the comparison of human and murine macrophages.
species grouped-label · level C · Concordant · 0 count
“Unlike in the case of WB comparisons, there were no enriched discordant modules in this comparison.”
Sample B — single-reviewer studies (10)
B1. Association between early arterial pH, base excess and lactate and 24-h mortality and neurological outcomes af
2026 · Resuscitation Plus · PubMed · DOI · study page · page 1 · adjudication: single reviewer
The area under the curve for lactate to predict 24-h mortality in rats was 0.959.
species rat · level A · Concordant · 0.959 auc
“In a multivariate regression analysis area under the curve, considering pH, base excess and lactate, for prediction of mortality were respectively: 0.796 (95%CI: 0.635-0.956), 0.980 (95%CI: 0.946-1.000), 0.959 (95%CI: 0.896-1.000) in rats”
B2. Translating metabolomic evidence gathered from an animal model to a real human scenario: the post-mortem inter
2025 · Metabolomics · PubMed · DOI · study page · page 7 · adjudication: single reviewer
Taurine and Hypoxanthine concentrations showed no differences between animal and human samples.
species grouped-label · level C · Concordant
“Interestingly, no differences in Taurine and Hypoxanthine concentrations can be seen between animal and humans as shown by Fig. 6.”
B3. Transcriptional programs diverge in aging mouse and human skeletal muscle
2026 · Aging · PubMed · DOI · study page · page 1 · adjudication: single reviewer
Neurogenesis demonstrated opposing or non-significant trends between mouse and human skeletal muscle with aging.
species grouped-label · level C · Mixed
“Hypoxia signaling, VEGFA, and inflammatory pathways showed concordant downregulation with aging in both species; however, angiogenesis, neurogenesis, and myogenesis demonstrated opposing or non-significant trends.”
B4. Transfer learning between preclinical models and human tumors identifies a conserved NK cell activation signat
2021 · Genome Medicine · PubMed · DOI · study page · page 10 · adjudication: single reviewer
The NK cell activation signature was significantly higher in anti-CTLA-4 responsive human tumors than non-responsive tumors.
species mouse · level C · Concordant · 1e-15
“In pre-treatment biopsies, the NK cell activation signature was significantly higher in anti-CTLA-4 responsive tumors than non-responsive tumors (p < 1 × 10-15, Additional file 1: Fig. S2A).”
B5. From signal to significance: characterizing anthracycline-related hypercoagulable adverse events via pharmacov
2026 · Frontiers in Pharmacology · PubMed · DOI · study page · page 1 · adjudication: single reviewer
The paper found that pathways related to cyclic adenosine monophosphate (cAMP) signaling and heparan sulfate biosynthesis were directionally consistent between experimental models and clinical samples.
species grouped-label · level C · Concordant
“Exploratory analyses suggested potential involvement of pathways related to cyclic adenosine monophosphate (cAMP) signaling and heparan sulfate biosynthesis, which were directionally consistent with observed alterations in coagulation-related parameters in clinical samples and experimental models.”
B6. Single-cell multimodal analysis identifies common regulatory programs in synovial fibroblasts of rheumatoid ar
2022 · Genome Medicine · PubMed · DOI · study page · page 20 · adjudication: single reviewer
17 shared regulons were identified between hTNFtg mice and human RA patients.
species grouped-label · level C · Concordant · 17 count
“A Regulatory network analysis in mouse and human datasets reveals 17 shared regulons.”
B7. Activin-A impedes the establishment of CD4+ T cell exhaustion and enhances anti-tumor immunity in the lung
2021 · Journal of Experimental & Clinical Cancer Research · PubMed · DOI · study page · page 1 · adjudication: single reviewer
Treatment of activin-A on tumor-infiltrating CD4+ T cells from lung cancer patients augmented their immunostimulatory capacity towards autologous CD4+ and CD8+ T cells, mimicking its anti-tumorigenic effects observed in the mouse lung cancer models.
species mouse · level C · Concordant
“Of translational importance, treatment of activin-A on tumor-infiltrating CD4+ T cells from lung cancer patients augmented their immunostimulatory capacity towards autologous CD4+ and CD8+ T cells.”
B8. Pan-glioma analyses reveal species- and tumor-specific regulation of neuron-glioma synapse genes by lncRNAs
2023 · Frontiers in Genetics · PubMed · DOI · study page · page 8 · adjudication: single reviewer
NGS genes in mouse PDX models showed similar expression levels to NGS genes in human MB.
species grouped-label · level A · Concordant
“Fourth, NGS genes in mouse MB PDX showed similar expression levels to NGS genes in human MB”
B9. Nanomedicine in 2026: Illustrative Quantitative Analyses of EPR Heterogeneity, Clinical Trial Attrition, and E
2026 · International Journal of Nanomedicine · PubMed · DOI · study page · page 6 · adjudication: single reviewer
Median nanoparticle tumour-to-blood ratio was 8.5 in murine models versus 1.8 in human tumours from the same publications (23 studies, 412 patients).
species mouse · level C · Discordant · 8.5 ratio
“compared to a median of 8.5 (IQR: 5.1-14.2) in murine models from the same publications.”
B10. Translatability of preclinical to early clinical tolerable and pharmacologically active dose ranges for centra
2023 · Translational Psychiatry · PubMed · DOI · study page · page 8 · adjudication: single reviewer
The preclinical data predicted the pharmacologically active range in humans to a high degree, with 18 out of 25 interventions showing an overlap of 80% or more for HED.
species grouped-label · level A · Concordant · 72 percent (n 18/25)
“Overall, the preclinical data predicted the pharmacologically active range in humans to a high degree, as indicated by an overlap of ≥ 80% in 18 out of 25 (72%) for HED, 15 out of 25 (60%) for Cmax and 19 out of 23 (83%) for AUC.”
Found something wrong? The item number, the question that failed and one line of what you saw is enough to act on — that is what the check is for.