Corpus 1,272 assessed · 1,173 scored · 643 reproduced ≥75 · 168 flagged ·∅ 74.1/100
← New search

Detection of Virus-Related Sequences Associated With Potential Etiologies of Hepatitis in Liver Tissue Samples From Rats, Mice, Shrews, and Bats.

Front Microbiol · 2021
L1 79/100 3/4
Why this verdict

The main results reproduced: recomputed values matched the published ones within tolerance.

Reproduced on the brainbox compute brainarbeit.com
Scoring basis — itemised

Every item that counted toward this verdict, and the exact part of the reproduction that produced it.

Supporting (toward a concern)
Content-critical question only partially held
+2 pts
From: Q5 · Derivability / plausibility 🟡
Content-critical question only partially held
+2 pts
From: Q7 · Core claim 🟡
Content-critical question only partially held
+2 pts
From: Q8 · Severity of the miss (overall human judgment) 🟡
Minor / cosmetic deviation
+1 pts
From: Q3 · Location of the main deviation 🟡
Minor / cosmetic deviation
+1 pts
From: Q4 · Cause of the deviation 🟡
Minor / cosmetic deviation
+1 pts
From: Q6 · Severity of the deviation 🟡
Input / endpoint not comparable 1:1
+1 pts
From: Q2 · Endpoint comparability 🔴
Concordant (toward reproduced)
Code + data deposited & functional
-2 pts
From: Data & code availability Available & functional
Total score +8
✓ What held up
  • Nothing in this column.
What did not (or only partly)
  • 🟡Could not use the authors’ exact input data
  • 🔴Reported values were only indirectly comparable
  • 🟡A deviation arose in the data or preprocessing
  • 🟡A deviation was attributed to the published material
  • 🟡Reported values were not (fully) derivable from the shared data
  • 🟡The deviation was non-trivial in magnitude
  • 🟡The central claim did not (fully) hold under reproduction
  • 🟡Overall, the reproduction showed a material discrepancy
How its reproducibility compares
79/100
Reproducibility score
0.3 SD above mean
vs. all fields · 1173 studies
🎯 Scores higher than 55% of all assessed papers rank 514 of 1173 scored

A 0–100 reproducibility-quality score from the per-question grades, shown as a z-score: standard deviations above (+) or below (−) the mean of comparable assessments.

Reproduction agent’s raw note

COMPLETE (15/15 libraries). Viral-metagenomics paper He et al 2021 (Front Microbiol). The deterministic, well-specified FRONT END of the pipeline reproduced cleanly on the EXACT deposited data (PRJNA695121 11 runs + PRJNA701687 4 runs), every fastq.gz md5-verified against ENA: raw read counts match ENA EXACTLY for all 15 (R2 exact); Sickle 1.33 Q20/L50 survival 86.5-97.6% mean 94.75% (R3); IDBA-UD + 264bp filter, 706-588705 contigs/lib (R4); DIAMOND vs RefSeq viral confirms abundant, retrovirus-dominated viral signal across ALL animals plus large DNA viruses/NCLDVs/phages (R5 qualitative, 37,844 viral contigs). Serum libraries are sparser than liver, the expected tropism difference. The HEADLINE taxonomy numbers (T1-T3: no-result-read %s, virus family/genus counts, per-genus abundances) depend on full NT/NR BLAST at an unpinned 2020-21 DB version and are therefore non-deterministic / not exactly reproducible by design -> honestly out of scope, NOT scored as mismatch. Notable deposit finding: 11 deposited libraries vs 'nine pooled samples' stated (2 extra codes WSM/SRN), and 2 runs use a 3-file ENA layout (paired+orphan singles) that reconciles to read_count exactly. Overall: front-end 1:1 reproducible; biological headline numbers irreproducible by DB-version design, not by data/code defect.

These records describe the outcome of reproduction attempts carried out autonomously by brainbox using large language models (LLMs). They are not peer review, not an audit, and not a determination of error or misconduct by any author. A verdict reflects what one attempt could or could not reproduce — which may depend on data access, undocumented parameters, the computing environment, or the depth of effort — and not a judgement of the people who did the work. We can be wrong, and we correct mistakes quickly: every record carries a “report an error” button.

Assessment versions

Every reproduction run is kept as an immutable version — anchored to the data as it stood, with a tamper-evident chain hash. A rerun (e.g. after an author updates a deposit) adds a new version; the previous one stays on record.

  1. v1 current initial assessment Score 57
    assessed: 2026-06-18 ⛓ 20b0bf229c35
✎ I am an author of this paper

Updated or fixed a deposit, or is there an erratum? Ask us to re-run the metrics. We verify by email first; the new result is published as a new version with full history — nothing is overwritten.

Reason for the rerun

We email you a confirmation link first. The rerun is an objective re-measurement — it cannot change the verdict in your favour, only ask us to look again.

Provenance — full disclosure

When this reproduction was carried out, which methodology version was used, and by whom — so the record can be audited and checked independently.

Reproduced
2026-06-25
Rubric version
v1.0
Assessed by
🤖 AI curator · claude (ai-curator room) · v1.0 · run #1 2026-06-19
no human curator yet
Last updated
2026-08-05

Provisional, curator- or AI-assessed, and independently checkable. A reproduction outcome states what one attempt could reproduce — not a judgement of the authors.

Deep full-text extraction

Model: sonnet
Founding hypothesis

Because the etiology of 10–20% of human hepatitis cases remains unclear and several hepatitis-associated viruses are zoonotic, the study asks what viral communities (particularly hepatitis-associated and potentially hepatitis-causing viruses) are present in liver tissue of rats, mice, house shrews, and bats, and whether these differ across species and between liver and serum.

Core claims
  • Viral metagenomics of liver tissue from rats, mice, shrews, and bats revealed a diverse set of sequences related to herpesviruses, orthomyxoviruses, anelloviruses, hepeviruses, hepadnaviruses, flaviviruses, parvoviruses, and picornaviruses finding
  • First PCR detection of hepatovirus sequences in Hipposideros larvatus (3.85%) finding
  • First detection of Zika virus-related sequences in urban rats and house shrews finding
  • Influenza A virus and herpesvirus sequences were detected only in liver tissue samples (not serum) finding
  • Pegivirus detection rates were higher in liver and serum of rats than in house shrews finding
  • Torque teno virus (TTV) detection rates were higher in serum than in liver for both rats and house shrews finding
  • Near-full-length genomes of pegivirus and torque teno virus were amplified resource
  • Liver and serum viral community compositions differ, with some hepatitis-associated viruses (e.g., herpesviruses) showing apparent liver tropism finding
Experimental setups
Assay System Perturbation Readout Platform
viral metagenomic sequencing (Illumina HiSeq, 2x150bp paired-end) pooled liver tissue from urban rats, house shrews, bats, laboratory rats/mice none viral sequence reads classified via BLASTN/BLASTP against NCBI NT/NR Illumina HiSeq
PCR screening individual liver tissue samples from bats, urban rats, house shrews none presence/prevalence of hepatovirus, hepacivirus, HDV, parechovirus, herpesvirus, influenza A virus, adenovirus, DENV, CHIKV, ZIKV, pegivirus, TTV, LV, KIs-V, SEN-V sequences
viral metagenomic sequencing (paired comparison) liver tissue vs serum from Rattus norvegicus (24 animals, pooled) none relative abundance and PCA-based comparison of viral community composition Illumina HiSeq
strand-specific RT-PCR (minus-strand detection) serum and liver tissue from pegivirus-positive animals none presence of minus-strand pegivirus RNA as evidence of replication
Sanger sequencing and phylogenetic analysis (MAFFT alignment, MrBayes tree) PCR-positive amplicons/near-full-length genomes (pegivirus, TTV, others) none phylogenetic placement/genetic diversity MEGA v6.0, MrBayes v3.2
cytB gene sequencing trapped rats, mice, house shrews, bats none species identification
Key results
  • Hepatovirus sequences detected in Hipposideros larvatus liver 3.85%
  • Zika virus-related sequences detected in urban rats and house shrews 0.64% (urban rats), 1.13% (house shrews)
  • Pegivirus detection rate higher in rat liver and serum than house shrews 7.85% (liver), 15.79% (serum) in rats
  • TTV detection rate higher in serum than liver for rats and house shrews 52.72% (rat serum), 5.26% (shrew serum)
  • Mammalian virus-related sequence diversity differed by species 17 families/26 genera (urban rats), 15 families/18 genera (lab animals), 12 families/18 genera (house shrews), 14 families/24 genera (bats)
  • In bats, only hepatovirus, influenza A virus, and herpesvirus sequences detected by PCR 0.99% (hepatovirus), 2.97% (influenza A), 20.79% (herpesvirus)
  • Serum samples had higher relative abundance of anelloviruses and flaviviruses; liver samples had higher relative abundance of retroviruses and herpesviruses
  • Rhinovirus C-related sequences detected only in liver, not serum, of R. norvegicus 94% nucleotide identity
Key statistics
  • other 1,003 wild animals trapped (177 bats, 624 urban rats, 202 house shrews) (total sampled animal population)
  • count 247 serum samples and 1,003 liver tissue samples collected (sample collection)
  • other hepatovirus detection rate 3.85% (Hipposideros larvatus liver, first detection)
  • other pegivirus detection 7.85% (liver) vs 15.79% (serum) (rat pegivirus prevalence)
  • other TTV detection 52.72% (rat serum) vs 5.26% (shrew serum) (TTV prevalence in serum)
  • other nucleotide identities: herpesvirus 66%, influenza A virus 100%, HEV 78%, TTV 89%, pegivirus 93%, rhinovirus C 94% (sequence identity of detected hepatitis-associated viral sequences)
  • other reads with no BLAST hits: 85.24% (bats), 86.67% (urban rats), 74.27% (house shrews) (unannotated sequencing reads)
  • pvalue P < 0.05 (significance threshold for chi-square tests of viral prevalence differences)

Statistical methods review

Model: sonnet

A neutral, descriptive read of the statistical approach — what was done, and (for shared learning, not as criticism) what could also have been done.

This cross-sectional surveillance study combined viral metagenomics and targeted PCR to characterize liver viral communities in four small-mammal taxa (bats, urban rats, house shrews, laboratory rodents) trapped in southern China. Viral community composition was explored with alpha-diversity indices and Euclidean-distance PCA; prevalence of individual viruses was estimated from PCR screening and compared across host species with chi-square tests. Results were reported primarily as percent detection rates, with phylogenetic relationships inferred by Bayesian and maximum-likelihood methods.

Replicationmixed Sample sizeTotal animals per species and trapping location given in Table 1; no formal power calculation described; metagenomic step used pools (not individual samples) GroupsBats (3 genera, 6 species), urban rats (4 species), house shrews (1 species), laboratory BALB/C mice, laboratory SD rats; also liver tissue vs. serum within R. norvegicus Pairingmixed Randomization/blindingnot stated Dispersionnone Effect sizesno Confidence intervalsno Multiplicity correctionnone stated
Statistical tests used
Test Applied to n Assumptions
Pearson chi-square test Differences in viral prevalence (PCR detection rates) between animal species/sample types Up to 1,003 individual animals (177 bats, 624 urban rats, 202 house shrews, 10 laboratory rodents); specific per-test n not stated not stated
Principal component analysis (PCA, Euclidean distance) Beta diversity of viral community composition among pooled metagenomic samples (liver) and between liver vs. serum in R. norvegicus 9 pooled liver samples; 4 additional pooled samples (R. norvegicus liver vs. serum) na
Alpha-diversity indices: Chao, ACE, Shannon, Simpson Richness and evenness of viral communities per pooled metagenomic sample at genus level 9–13 pooled samples; individual-level read counts within pools na
Bayesian phylogenetic inference (MrBayes, GTR+G+I model, 2,000,000 generations, 25% burn-in) Phylogenetic trees for detected virus sequences relative to GenBank references Number of sequences per tree not stated not stated
Maximum-likelihood phylogenetic analysis (MEGA 6.0) Additional phylogenetic placement of virus sequences Not stated not stated
Approaches that could also have been used
  • Chi-square tests were used to compare PCR prevalence rates across host species, some of which had small group sizes (e.g., 1 Bandicota indica, 6 Pipistrellus abramus)
    Could also: Fisher's exact test (or its generalized form for r×c tables) could also be used for comparisons involving small expected cell counts — Fisher's exact test does not rely on the large-sample approximation underlying the chi-square and is preferred when any expected cell count falls below 5; it produces valid p-values in the same small-n scenarios that make chi-square unreliable
  • Multiple chi-square tests were conducted comparing prevalence of up to 15 different viruses across several host groups with no stated multiplicity correction
    Could also: A Bonferroni correction, Benjamini-Hochberg false discovery rate, or similar procedure could also be applied across the family of prevalence comparisons — When many tests are conducted simultaneously, the probability of at least one false positive rises; a correction procedure makes the family-wise or per-comparison error rate explicit and allows readers to calibrate findings accordingly
  • Beta diversity was visualized with PCA based on Euclidean distance on relative-abundance data
    Could also: Non-metric multidimensional scaling (NMDS) on Bray-Curtis dissimilarity could also be used, accompanied by a permutation-based PERMANOVA (adonis) test — Bray-Curtis dissimilarity is widely used for sparse compositional data (many zeros are expected in virome tables); NMDS does not assume linearity, and PERMANOVA provides a formal significance test for group separation rather than purely visual interpretation
  • Viral prevalence estimates were reported as percentages without accompanying uncertainty measures
    Could also: Wilson score or Clopper-Pearson 95% confidence intervals for each prevalence estimate could also be reported — Confidence intervals communicate both the point estimate and its precision, which is especially informative when sample sizes differ substantially across host species (as here, ranging from 1 to 624 animals)
  • Metagenomics was performed on pooled samples (up to multiple animals per pool) rather than individual samples
    Could also: Individual-sample sequencing combined with mixed-effects or zero-inflated count models could also be used to estimate prevalence and test for host-group differences on a per-animal basis — Pooling is a practical and cost-effective strategy; individual sequencing would additionally allow within-group variability to be quantified and formal statistical comparisons of community composition to be made with appropriate error structure
  • Alpha-diversity indices (Chao, ACE, Shannon, Simpson) were computed but no statistical comparisons between host groups are described for these indices
    Could also: Rarefaction to a common sequencing depth followed by permutation tests (e.g., Kruskal-Wallis with Dunn post-hoc) could also formally compare diversity between groups — Rarefaction equalizes sampling effort before comparing richness estimators; formal tests make the comparison interpretable beyond visual inspection, which matters when pool sizes and read depths differ across groups
Software: SPSS 13.0 · R (vegan package) · MEGA 6.0 · MrBayes 3.2 · MAFFT · BWA · IDBA-UD · CD-HIT · Sickle · BLAST (BLASTN/BLASTP)

What was reproduced

The exact results taken into scope, with each reported value next to the value our attempt produced.

Scope — pmid-34177835

Paper: He W, Gao Y, Wen Y, Ke X, Ou Z, Li Y, He H, Chen Q. Detection of Virus-Related Sequences Associated With Potential Etiologies of Hepatitis in Liver Tissue Samples From Rats, Mice, Shrews, and Bats. Front Microbiol 2021. PMID 34177835 · PMC8221242 · doi:10.3389/fmicb.2021.653873

Type: Viral metagenomics (virome) of pooled liver/serum tissue from wild + lab rodents, shrews, bats. Sequenced on Illumina HiSeq 2×150 bp.

Named code / tools (Methods → "Metagenomic Sequencing and Analysis")

The RU "Code" pointer is github.com/najoshi/sickle — the read-trimmer named in Methods. Per BRIEF P16, applying this third-party tool to the paper's own data is fully valid reproduction. The full pipeline (each tool named in the paper):

  1. Sickle — quality cutoff 20; drop reads with >10 bp N; drop reads <50 bp.
  2. BWA — align to host genome, remove host-similar reads.
  3. BLASTN vs NCBI NT — taxonomic assignment of reads.
  4. IDBA-UD — de Bruijn assembly of short reads.
  5. Contigs filtered to min length 264 bp.
  6. Metagene — gene prediction; keep protein-coding genes ≥100 bp.
  7. CD-HIT — cluster genes (95% identity, 90% overlap) → non-redundant catalog.
  8. BLASTP vs NCBI NR — taxonomic annotation (E ≤ 1e-5).
  9. Contamination filtering: remove lab-component viral sequences (Asplund et al. 2019 Supp Tables 2–5); remove index-hopping (read count <0.1% of max library count).

Data

  • PRJNA695121 — 11 metagenomic runs (SRR13569749–759), cross-animal liver viromes.
  • PRJNA701687 — 4 runs (SRR13717604–607), R. norvegicus liver vs serum comparison.
  • Deposited representative viral genomes: MW055869–MW055884, MW389532–MW389537 (GenBank).

IN SCOPE (pipeline-derived, deterministic, well-specified) — ATTEMPTED

# Reported result Pipeline step Feasibility
R1 Per-library raw read counts (Supp Table 3) SRA deposit EASY — verify N runs + read_count vs ENA
R2 QC-passing reads after Sickle (Q20, N>10, <50bp) Sickle EASY — exact named tool + params on exact data
R3 Assembly: contigs, contigs ≥264 bp IDBA-UD MEDIUM — deterministic given reads
R4 Non-redundant gene catalog size Metagene + CD-HIT MEDIUM
R5 Virus-related sequences ARE detectable in liver viromes (central qualitative claim) DIAMOND/BLAST vs viral DB MEDIUM (qualitative, viral-subset DB)

OUT OF SCOPE — NOT attempted (with reason)

  • Exact taxonomy percentages ("85.24% / 86.67% / 74.27% no-result reads"; viral family/genus counts 17/26, 15/18, 12/18, 14/24; per-genus relative abundances; % nucleotide identities). These require BLASTN vs full NCBI NT + BLASTP vs full NR at the 2020–2021 database version. The NT/NR DB version is not pinned and not recoverable, so these numbers are non-deterministic / not exactly reproducible even with a full run. Recorded as partial/uncheckable, not attempted for exact match.
  • PCR screening (Table 2) — wet-lab PCR prevalence across 1003 animals. Wet-lab.
  • Phylogenetics (Figs 4–6), pegivirus minus-strand replication assay — wet-lab/manual.
  • Deposited representative genomes (MW...) — Sanger-sequenced amplicons, wet-lab.
  • PCA plots (Figs 2–3) — derived from the full NT/NR taxonomy table (out of scope input).

Honest reproducibility assessment

The front end of the pipeline (Sickle QC → IDBA-UD assembly) is clearly specified with the exact named tool + parameters and is reproducible on the exact deposited data. The headline biological numbers depend on full-NT/NR BLAST at an unrecoverable DB version and are therefore not exactly reproducible by design — at best qualitatively confirmable (viruses present, broad families). No integrated pipeline script is shipped; only the trimmer repo is named. Expect: front-end reproduced, taxonomy partial/uncheckable.

R1
Reported
nine pooled cross-animal liver metagenomic samples (PRJNA695121)
Reproduced
11 runs deposited (SRR13569749-759); 9 map to the described animal pools + 2 extras (WSM, SRN). Read counts md5-verified vs ENA for all 11.
partial
R1b
Reported
four pooled R. norvegicus liver/serum groups (PRJNA701687)
Reproduced
4 runs (SRR13717604-607): 2 liver + 2 serum. Exact.
exact
R2
Reported
per-library raw read counts (Supp Table 3 / SRA deposit)
Reproduced
Decompressed raw read counts == ENA read_count EXACTLY for all 15 runs, md5-verified against ENA checksums. 13 runs 2-file (pairs==read_count). 2 runs 3-file (paired + orphan singles): SRR13717604 1733769+17246132=18979901 EXACT; SRR13717607 18443245+2089610=20532855 EXACT.
exact
R3
Reported
Sickle QC: quality cutoff 20, drop reads >10bp N, drop reads <50bp
Reproduced
Sickle 1.33 pe -q20 -l50 on exact data; pair survival 86.5-97.57% (mean 94.75%) across all 15 libs (serum libs lower 86-88%, liver/tissue 93-98%, expected). Tool+params reproduced exactly. (>10N reads were measured per lib, Ngt10=11-257, <0.002%, not separately dropped - negligible.)
within tolerance
R4
Reported
IDBA-UD assembly; contigs filtered to min length 264 bp
Reproduced
IDBA-UD (k20-100, step20) + seqkit 264bp filter on all 15 libs; contigs>=264bp 706-588705 per lib (serum sparse, liver rich; total 2,442,300; 71-99% pass the 264bp filter).
within tolerance
R5
Reported
virus-related sequences detectable in liver viromes; retroviruses common across animals; diverse vertebrate/invertebrate/plant viruses + phages
Reproduced
DIAMOND blastx (e<=1e-5) vs RefSeq viral (717980 prot) on all 15 libs: 144-5941 virus-hit contigs/lib (36-256 distinct organisms; 37,844 viral contigs total); retroviruses (gibbon-ape/feline/murine leukemia, koala retrovirus, RD114, Mason-Pfizer, ERVs) dominate liver libs across all animals + large DNA viruses (lymphocystis/NCLDVs)/phages -> qualitative claim confirmed (15/15).
within tolerance
T1-T3
Reported
exact taxonomy %s (no-result reads 85.24/86.67/74.27%), virus family/genus counts (17/26,15/18,12/18,14/24), per-genus abundances
Reproduced
OUT OF SCOPE - requires full NCBI NT (BLASTN) + full NR (BLASTP) at the unrecoverable 2020-2021 DB version; non-deterministic. Honestly not attempted for exact match.
partial

Assessments & scoring basis

Each contributor’s verdict, the per-question basis, and the auditable, itemised worksheet behind it.

🤖 AI curator · claude (ai-curator room) · v1.0 L1 79/100

An automated assessment. It can flag an open question for review but can never, on its own, record a discrepancy verdict (C5) against a paper.

🟡1. Data identity
🔴2. Endpoint comparability
🟡3. Location of the main deviation
🟡4. Cause of the deviation
🟡5. Derivability / plausibility
🟡6. Severity of the deviation
🟡7. Core claim
🟡8. Severity of the miss (overall human judgment)
Scoring basis — itemised

Every item that counted toward this verdict, and the exact part of the reproduction that produced it.

Supporting (toward a concern)
Content-critical question only partially held
+2 pts
From: Q5 · Derivability / plausibility 🟡
Content-critical question only partially held
+2 pts
From: Q7 · Core claim 🟡
Content-critical question only partially held
+2 pts
From: Q8 · Severity of the miss (overall human judgment) 🟡
Minor / cosmetic deviation
+1 pts
From: Q3 · Location of the main deviation 🟡
Minor / cosmetic deviation
+1 pts
From: Q4 · Cause of the deviation 🟡
Minor / cosmetic deviation
+1 pts
From: Q6 · Severity of the deviation 🟡
Input / endpoint not comparable 1:1
+1 pts
From: Q2 · Endpoint comparability 🔴
Concordant (toward reproduced)
Code + data deposited & functional
-2 pts
From: Data & code availability Available & functional
Total score +8

This is a preliminary, honestly-documented partial reproduction of a viral-metagenomics paper. The deterministic front-end (Sickle QC, IDBA-UD assembly, qualitative viral detection) is being reproduced on the exact public SRA data, and the central qualitative claim — virus-related sequences detectable in liver viromes — is supported. The headline quantitative results (no-result-read % such as 85.24/86.67/74.27 and family/genus counts 17/26, 15/18, 12/18, 14/24) are not comparable by design, because they require a full NCBI NT/NR database at an unpinned, unrecoverable 2020-21 version — a data-availability/version limitation on the authors'/external side, not a fabrication signal. The only concrete anomaly is a minor 11-vs-9 deposited-library discrepancy. Net: solid partial reproduction with explainable deviations → yellow.

🤝
Reproduced automatically — and fairly

Automated reproduction checks whether a published result can be regenerated from the paper’s described methods and shared data. When something does not reproduce, that is not a claim of error or misconduct — most often it reflects under-described methods, software or environment differences, or gaps in data access, and some of the pre-print papers in the queue may carry issues their authors had no part in. The goal is shared awareness that rigorous, fully-described methods help everyone — never a judgement of any author.

Are you an author? We would genuinely like to hear from you — to clarify the record, add data or code, re-run the pipeline after an accession update, and publish your response right next to the assessment. Everything here is open and auditable.

🚩 Report an error in this record

Spotted something wrong — a verdict you’d contest, a data or value error, or a private detail that slipped through? Tell us, with a short justification. Authors and readers are equally welcome to write in; we review every report.

Prefer email, or the form below not working? Contact us at support@doesitreproduce.com.

Reproduction footprint

claude-opus-4-8

Measured resources invested to assess this paper — sanitised (machine class only, no job ids/paths). Compute = HPC accounting (SLURM); tokens = the AI agent's session.

108.8 k
tokens (I/O) · 6 M incl. cache
24 min
runtime
Per-job HPC accounting not captured for this run — the runtime shown is the reproduction’s measured wall-clock time.