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

A Deeper Insight into Evolutionary Patterns and Phylogenetic History of ASFV Epidemics in Sardinia (Italy) through Extensive Genomic Sequencing.

Viruses · 2021
L1 76/100 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 🟡
Minor / cosmetic deviation
+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
76/100
Reproducibility score
at the mean
vs. all fields · 1173 studies
🎯 Scores higher than 48% of all assessed papers rank 586 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

PARTIAL (honest, data-availability-limited). «our HPC» reachable; SLURM «job» ran the full per-sample pipeline on n094. Pipeline: ENA fastq -> Trim Galore --paired -> bwa-mem vs ASFV ref KX354450 (strain 47/Ss/2008, Stintino/Sardinia, genotype I, 184,638 bp -- CORRECTED: scope.md had mislabeled it Georgia 2007/1) -> samtools sort -> Picard MarkDuplicates -> FreeBayes(--ploidy 1 -X -u -m 20 -q 20 -F 0.2) -> bcftools consensus (QUAL>=20, zero-cov masked). Single SRA run SRR13976567 (isolate NU1981_2, 68,097 read pairs, 99.98% mapped). RESULT: 3/4 per-sample metrics land inside the paper's reported cohort ranges -- length 181,640 bp non-N (reported 181,684-181,925), GC 38.60% (reported 38.56-38.60), median depth 80-82x (reported 6-250x). C4 (95 'new' mutations) is a cohort aggregate over all 71 genomes and is NOT reproducible from one genome; per-sample freebayes with the paper's exact params yields 42 confident SNPs vs KX354450. Described well enough to reproduce the per-sample pipeline 1:1. NOT attempted: cohort phylogenetics/dating and the 58-isolate cohort assembly (only 1 of 71 genomes provided). DATA NOTE: NCBI sra-tools fasterq-dump de-paired this run (R1 != R2); ENA direct fastq gave clean pairs.

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 50
    assessed: 2026-06-18 ⛓ 3db117afccb7
✎ 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

The study tests whether genetic variability within a single, spatially-localized ASFV outbreak can be adequately captured by a small number of whole genomes, or whether a much larger sequencing effort improves phylogenetic resolution and evolutionary/epidemiological inference of the Sardinian ASFV epidemic.

Core claims
  • 58 new whole genomes of Sardinian ASFV isolates were sequenced, the largest ASFV whole-genome sequencing effort to date resource
  • The new sequences capture more than twice the genomic and phylogenetic diversity of all previously published Sardinian ASFV sequences finding
  • The increased diversity enables, for the first time, resolution of genetic substructure within the outbreak, revealing multiple ASFV subclusters, some coexisting in space and time finding
  • This is the first phylodynamic inference of ASFV based on whole genomes applied to a closed epidemic system method
  • Sardinian ASFV isolates after 1990 (modern strains) show deletions in both the B602L and EP402R genes compared to historical strains isolated before 1990 finding
  • Ninety-five new point mutations were detected, located within MGF360, MGF110 and MGF505 gene families, along with seven URFs and indels finding
  • Genome annotation identified 231 ORFs (165 protein-coding genes and 66 uncharacterized reading frames) finding
Experimental setups
Assay System Perturbation Readout Platform
Whole-genome sequencing (NGS) ASFV isolates from porcine monocyte/macrophage cultures (domestic pig, wild boar, illegal free-ranging pig samples, Sardinia 1978-2018) none full viral genome sequence, nucleotide mutations, indels Illumina HiSeq 2500 and NovaSeq 6000, Nextera XT/DNA Flex library prep
Sanger sequencing B602L and EP402R gene regions of ASFV isolates none confirmation of gene sequence/deletions Sanger sequencing
Haemoadsorption test (Malmquist test) porcine monocyte/macrophage monolayers none presence of infectious ASFV
Bayesian phylogenetic reconstruction 71 Sardinian ASFV WGSs plus 3 outgroup genomes (L60, E75, Benin97) none tree topology, clustering/subgroup structure MrBayes 3.2.7
Molecular dating / phylodynamic analysis same 71-genome WGS dataset none time-scaled MCC tree, evolutionary substitution rate, lineage-through-time dynamics BEAST 1.8.2, Tracer 1.7
Principal coordinate analysis (PCoA) ASFV genomes from 2002-2008 cluster (n=46) none genetic dissimilarity/subgroup structure GenAlEx 6.5
Likelihood-mapping analysis genome alignment of ASFV dataset none phylogenetic signal reliability (noise/saturation/recombination check) TreePuzzle
Pearson's Chi-squared test phylogenetic clusters vs. geographic origin of isolates none statistical association between genetic structuring and geography R stats package
Key results
  • 58 new ASFV whole genomes sequenced, representing the largest ASFV WGS effort to date
  • New sequences capture substantially more genomic and phylogenetic diversity than all previously published Sardinian sequences combined more than twice
  • Multiple ASFV subclusters identified within the Sardinian epidemic phylogeny, some coexisting in space and time
  • Complete genome sequences obtained for a subset of the 58 isolates analyzed 40 of 58
  • New point mutations detected within MGF360, MGF110 and MGF505 gene families 95
  • Genome annotation yielded protein-coding genes and uncharacterized reading frames 231 ORFs (165 protein-coding, 66 URFs)
  • Sequenced genome lengths fell within a narrow range 181,684–181,925 bp
  • GC content of sequenced genomes was consistent across isolates 38.56–38.60%
Key statistics
  • count 58 (new ASFV whole genomes sequenced in this study)
  • count 40/58 (isolates with complete genome sequences obtained)
  • other 181,684–181,925 bp (range of assembled genome lengths)
  • other 38.56–38.60% (range of GC content across sequenced genomes)
  • count 231 ORFs (165 protein-coding, 66 URFs) (GATU genome annotation results)
  • count 95 (new point mutations detected across MGF360, MGF110, MGF505)
  • count 2253 (total ASFV outbreaks reported in Sardinia, 1978-2018)
  • count n=46 (genomes from 2002-2008 cluster analyzed by PCoA)

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 phylogenomic study characterized the evolutionary history of ASFV in Sardinia (1978–2018) using 74 whole-genome sequences (58 newly generated, 13 published, plus 3 outgroup genomes). Phylogenetic structure was inferred via Bayesian MCMC (MrBayes) and molecular dating via Bayesian phylodynamics (BEAST), with clock and demographic model selection by Bayes Factor comparison. Genetic diversity was quantified using Faith's phylogenetic diversity index and pairwise nucleotide statistics, and a Pearson's chi-squared test with Monte Carlo permutation was used to test geographic–phylogenetic association.

Replicationbiological Sample size58 new isolates selected (~15% of ~400 archived strains) based on spatial and temporal distribution of outbreaks; at least 5% of samples included from each epidemic year; no formal power calculation stated GroupsSardinian ASFV isolates across sampling years (1978–2018), host species (domestic pig, wild boar, illegal free-ranging pig), and geographic province; three outgroup genotype I sequences from Africa and Iberian Peninsula included Pairingna Randomization/blindingnot stated Dispersionnone Effect sizesno Multiplicity correctionnone stated
Statistical tests used
Test Applied to n Assumptions
Bayesian MCMC phylogenetic inference (GTR+I+G model, two independent runs of 4 Metropolis-Coupled chains, 5 million generations, 25% burn-in; convergence via ASDSF and PSRF) Main phylogeny of 74 ASFV whole genomes (71 Sardinian + 3 outgroups) 74 not stated
Bayesian molecular dating with strict and uncorrelated log-normal relaxed clock models under constant, exponential, expansion, and Bayesian Skyline demographic models; clock/demographic model selected by Bayes Factor (2lnBF via Tracer); final MCMC 100 million generations, ESS >200 required Estimation of evolutionary rate and time-calibrated phylogeny of Sardinian ASFV; also applied to subsets to investigate temporal changes in genetic diversity 71 not stated
Likelihood mapping of 10,000 random quartets (TreePuzzle) Verification of phylogenetic signal in the whole-genome alignment before phylodynamic inference 74 not stated
Faith's phylogenetic diversity index on neighbor-joining trees (Hamming distance per base); quantiles from 100 random subsamples of each size Quantification of phylogenetic information contributed by additional sequences across different subsample sizes Up to 71 Sardinian sequences (subsamples of varying size) not stated
Principal coordinate analysis (PCoA, GenAlEx 6.5) based on genetic dissimilarity among genomes Identification of virus subgroups within the 2002–2008 genetic cluster 46 not stated
Pearson's chi-squared test with Monte Carlo simulation (2000 replicates; R stats package) Test of association between phylogenetic cluster membership and geographic origin of isolates not stated
Approaches that could also have been used
  • Clock and demographic model selection was performed using the Bayes Factor (2lnBF) estimated from MCMC log-likelihoods in Tracer, which corresponds to the harmonic mean estimator
    Could also: Stepping-stone or path-sampling marginal likelihood estimation in BEAST could also be used for model comparison — Stepping-stone and path-sampling are considered more numerically stable estimators of the marginal likelihood than the harmonic mean; they are increasingly recommended in phylodynamic workflows when computing 2lnBF for clock and demographic model selection
  • The main phylogeny was inferred using Bayesian MCMC with MrBayes under GTR+I+G
    Could also: Maximum likelihood phylogeny with ultrafast bootstrap support (e.g., IQ-TREE 2) could also have been constructed as a complementary or independent tree — ML inference is computationally faster and produces bootstrap support values that are straightforward to interpret; it is widely used as a cross-check on Bayesian topologies, especially for large genome alignments
  • Phylogeographic association between cluster membership and geographic origin was assessed with a Pearson's chi-squared test (with Monte Carlo permutation)
    Could also: Bayesian Tip Significance testing (BaTS) or a Mantel test on patristic genetic distances versus geographic distances could also be applied — BaTS explicitly accounts for phylogenetic uncertainty across the posterior tree distribution when testing trait–phylogeny associations, and a Mantel test evaluates continuous isolation-by-distance; either approach would complement a chi-squared test that treats cluster assignment as discrete and fixed
  • Subgroup structure in the 2002–2008 cluster was visualized using PCoA in GenAlEx
    Could also: A neighbor-net phylogenetic network (e.g., SplitsTree4) or PCA directly on the SNP matrix could also depict genetic structure — Network methods represent reticulate relationships and homoplasy more explicitly than distance-based PCoA, which can be particularly informative for dense, low-divergence viral genomes where tree-like signal may be incomplete
  • Clock signal was assessed indirectly through BEAST model comparison and ESS diagnostics rather than a dedicated pre-analysis check
    Could also: A root-to-tip regression against sampling date (e.g., in TempEst) prior to BEAST analysis could also be used to evaluate clock signal and flag temporal outliers — Root-to-tip regression explicitly tests whether collection dates correlate with genetic divergence — a prerequisite for reliable Bayesian molecular dating — and can identify sequences that are inconsistent with a molecular clock before investing substantial computational resources
  • The sequencing effort (n = 58 new isolates, ~15% of archived strains, at least 5% per year) was determined by pragmatic sampling criteria without a formal power analysis
    Could also: Simulation-based power analyses for Bayesian phylodynamic inference could also be used to estimate how many sequences are needed to recover reliable evolutionary rate and demographic history estimates — Because the paper explicitly poses the question of how many sequences suffice for reliable phylodynamic inference, a simulation framework (e.g., using BEAST with synthetic data of varying n) would provide a quantitative answer alongside the empirical rarefaction analysis already performed
Software: R/ape R 4.1.0 · MrBayes 3.2.7 · BEAST 1.8.2 · Tracer 1.7 · JModelTest 2.1.7 · MAFFT 7.427 · GenAlEx 6.5 · TreePuzzle · FigTree 1.4.0 · TreeAnnotator (BEAST package) · CIPRES Science Gateway · Jalview 2.10.3B.1 · freebayes · GATU

What was reproduced

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

Scope — pmid-34696424

Paper: Fiori et al. 2021, Viruses 13:1994. "A Deeper Insight into Evolutionary Patterns and Phylogenetic History of ASFV Epidemics in Sardinia (Italy) through Extensive Genomic Sequencing." DOI 10.3390/v13101994 · PMCID PMC8539718.

Code link in record: https://github.com/FelixKrueger/TrimGalore (third-party read-trimming tool — valid per brief rule P16: applying an existing third-party tool to the paper's own data is a legitimate reproduction).

Data handed to this room: sra:SRR13976567 — a SINGLE run = sample NU1981_2 (BioSample SAMN18312889), ASFV WGS, Illumina NovaSeq 6000, paired-end 2×150, 68,156 read pairs / 20.06 Mbp. This is ONE of the 57/58 newly sequenced isolates.

Pipeline described (Methods 2.3, 2.5)

  1. Trim Galore — quality-trim + adapter removal of raw reads.
  2. bwa-mem — align reads; reads mapping uniquely to the ASFV genome retained. (Reference accession KX354450 = ASFV strain 47/Ss/2008 (Stintino, Sardinia, genotype I, 184,638 bp) — CORRECTED 2026-06-25: an earlier draft of this scope mislabeled KX354450 as "Georgia 2007/1"; that is wrong, KX354450 is the Sardinian genotype-I reference, which is why the new isolates' lengths (~181.6 kb) sit just below it. Verified against PMC8539718 Methods + GenBank.) Note: Methods text mentions Sus scrofa 10.2 to subtract host; the ASFV consensus is built against KX354450.
  3. samtools sort+index; Picard MarkDuplicates (dedup).
  4. GEM re-alignment (host/uniqueness filtering).
  5. Freebayes variant calling: --ploidy 1 -X -u -m 20 -q 20 -F 0.2.
  6. Consensus genome per sample → MSA → phylogenetics (MrBayes 3.2.7: NST=6, rates=invgamma, ngammacat=4; BEAST 1.8.2 molecular dating).

IN SCOPE (pipeline-derived, attempted on SRR13976567 = NU1981_2)

# Reported result Pipeline Note
C1 ASFV genome length per isolate: 181,684–181,925 bp (range over all) TrimGalore→bwa-mem→consensus vs KX354450 check NU1981_2 length within/near range
C2 GC content 38.56–38.60% consensus check NU1981_2 GC in range
C3 Median coverage 6–250× (Table S2) bwa-mem depth check NU1981_2 depth plausible
C4 Point mutations vs reference (95 new across dataset; Table S3) Freebayes count high-confidence variants in NU1981_2 vs KX354450

OUT OF SCOPE (not attempted)

  • Wet-lab: DNA extraction, library prep, the actual Illumina sequencing.
  • Aggregate / multi-genome results: the 95 new point mutations, MrBayes tree topology, BEAST tMRCA / molecular dating, recombination, selection analyses — these require ALL 71 WGS genomes (only 1 run was provided to this room) and external reference panels. Out of scope: single-accession room.
  • N reported = 58 new isolates / 71 total: only 1 SRA run available here; we profile that one run, not the cohort.

Honesty note

Single-run reproduction: we can faithfully run the per-sample read→consensus→variant pipeline on NU1981_2 and check the derived metrics fall in the paper's reported ranges. We CANNOT reproduce cohort-level phylogenetics from one accession. This is a genuine partial reproduction by data availability, not a failure.

Figures / tables: Table
C1
Reported
ASFV consensus genome length 181,684-181,925 bp (cohort range)
Reproduced
181,640 bp non-N (184,650 bp total incl 3,010 N), NU1981_2 vs KX354450
within tolerance
C2
Reported
GC content 38.56-38.60%
Reproduced
38.60%
within tolerance
C3
Reported
median coverage between 250 and 6x (cohort range, Table S2)
Reproduced
median 80x whole-genome / 82x covered; mean 93x; 98.37% breadth
within tolerance
C4
Reported
95 new point mutations (cohort aggregate over 71 genomes)
Reproduced
42 SNPs QUAL>=20 (+16 indels) for NU1981_2 vs KX354450; freebayes paper params
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 76/100

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 🟡
Minor / cosmetic deviation
+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 an incomplete, in-progress reproduction: the pipeline (TrimGalore→bwa-mem vs KX354450→samtools/Picard→Freebayes) is scoped and scripted, reported cohort values (genome 181,684–181,925 bp; GC 38.56–38.60%; coverage 6–250x; 95 new mutations) are recorded, but no reproduced values exist because «our HPC» compute is pending and compare.py has not run. The only deviation so far is on our/data-availability side: 1 of 58 SRA runs was handed to the room, so cohort-level and phylogenetic claims are out of scope rather than refuted. There is no fabrication signal — the underlying SRA data is public and the reported per-isolate ranges are plausibly derivable once the single run completes. Graded all-yellow: a solid setup with explainable scoping/coverage deviations and nothing yet computed to mark green or red.

🤝
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.

44.6 k
tokens (I/O) · 2.1 M incl. cache
20 min
runtime
Per-job HPC accounting not captured for this run — the runtime shown is the reproduction’s measured wall-clock time.