Analysis of the genome of the New Zealand giant collembolan (Holacanthella duospinosa) sheds light on hexapod evolution.
The main results reproduced, with only marginal, non-material deviations.
Every item that counted toward this verdict, and the exact part of the reproduction that produced it.
- ✓Reported values were directly comparable
- ✓No authors-side cause for any deviation
- ✓Reported values are derivable from the shared data
- ✓The central claim held under reproduction
- 🟡Could not use the authors’ exact input data
- 🟡A deviation arose in the data or preprocessing
- 🟡The deviation was non-trivial in magnitude
- 🟡Overall, the reproduction showed a material discrepancy
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
DESCRIBED WELL ENOUGH -> reproduced 1:1 on the robust metrics; partial on size/N50 for a fully-explained reason. Wu et al. 2017 (BMC Genomics) assembled the Holacanthella duospinosa (NZ giant springtail) genome from ~72 Gb Illumina (PRJNA384703) with SOAPdenovo2/GapCloser/SSPACE. We did NOT re-assemble from raw reads (the heavy 20%); instead we applied standard third-party QC tools (assembly-stats, seqkit, BUSCO v5.7.1) to the paper's own deposited assembly GCA_002738285.1 (ASM273828v1) -- per brief P16 an equally valid reproduction. KEY HONEST FINDING: the public deposit is a FILTERED version of the paper's Table-2 assembly (NCBI drops <200 bp scaffolds [shortest=200] + flagged contaminants), so it has 62,430 seqs / 327.6 Mbp vs the paper's 410,937 seqs / 370.3 Mbp, and N50 is correspondingly higher (310,226 vs 226,503). Metrics robust to that filtering reproduce cleanly: longest scaffold is BIT-IDENTICAL (2,807,427 bp), GC 33.35% vs 33.40%, N 2.49% vs 2.18%; and our recompute matches NCBI's OWN published assembly stats (327.6 Mbp / N50 310,226 / GC 33.5) exactly, validating the computation. BUSCO completeness C:95.4% essentially equals the paper's reported 95.3% complete (the paper's stricter 77.4% 'full-length' reflects the older BUSCO v2.0.1/odb9/AUGUSTUS method; we used v5.7.1/odb10/miniprot). NO fabrication signal: every reported number is consistent with the deposited data once NCBI filtering and tool-version differences are accounted for. NOT ATTEMPTED (hard 20%, by design): de-novo re-assembly from SRA reads; repeat content 42.96% (RepeatModeler+RepeatMasker, many CPU-h); gene annotation 12,000/9,911 models (MAKER2+Augustus+FGENESH); Trinity transcriptome; HGT counts (manual); flow-cytometry genome size 320 Mbp (wet-lab, out of scope). Grades: 1 exact, 3 within-tol, 3 partial, 0 mismatch.
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.
-
v1 current initial assessment Score 72assessed: 2026-06-16 ⛓ 45e1ced12fbd
✎ 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.
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
- 2026-06-16
- Rubric version
- v1.0
- Assessed by
-
🤖 AI curator · claude (ai-curator room) · v1.0 · run #1 2026-06-16no 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: opusWhat does the draft genome of the New Zealand giant collembolan Holacanthella duospinosa reveal about the genetic basis of key hexapod traits (sex determination, DNA methylation, chemoreception, chitinase activity), its phylogenetic placement, and horizontal gene transfer within Collembola?
- ★ A high-quality ~375 Mbp draft genome and transcriptome of Holacanthella duospinosa was assembled and annotated, providing a genomic resource for hexapod evolution. resource
- ★ Phylogenomic analysis places H. duospinosa within the Neanuridae. finding
- ★ H. duospinosa lacks an intact DNA methylation system, shown by unimodal CpG[o/e] distribution and absence of Dnmt1 and Dnmt3 orthologs. finding
- ★ The species lacks odorant receptors and their obligate co-receptor Orco, while retaining ionotropic glutamate and gustatory receptors. finding
- ★ Horizontal gene transfer of bacterial and fungal genes occurs in H. duospinosa, confirming HGT as a general phenomenon across Collembola. finding
- ★ Orthologs of conserved sex determination genes were recovered in the genome. finding
- Twenty-three chitinase-like genes were identified, potentially involved in digesting fungal cell walls in this saproxylic organism. finding
- The SSU rRNA gene contains a 1454 bp insertion encoding a reverse transcriptase, suggesting a retrotransposon within the rRNA. finding
| Assay | System | Perturbation | Readout | Platform |
|---|---|---|---|---|
| Whole-genome DNA sequencing and de novo assembly | Holacanthella duospinosa (collembolan) | none | Genome assembly size, N50, coverage, GC content, heterozygosity | Illumina |
| Transcriptome (RNA-seq) sequencing and assembly | Holacanthella duospinosa | none | Contig number, N50, length distribution | Illumina |
| Genome completeness assessment | H. duospinosa assembly | none | Percent complete/partial conserved genes | BUSCO v2.0.1 (arthropoda_odb9) |
| Repeat/transposable element annotation | H. duospinosa genome | none | Repeat class proportions and copy numbers | — |
| Structural gene annotation and homology BLAST | H. duospinosa genome/proteome | none | Gene models, protein homology to NCBI nr | BLASTP (e-value 1e-6); TBLASTN |
| Phylogenomic analysis | H. duospinosa vs other arthropods | none | Phylogenetic placement (370,877 amino acids) | — |
| DNA methylation inference via CpG[o/e] content | H. duospinosa, Apis mellifera, Drosophila melanogaster gene bodies/transcripts | none | CpG[o/e] distribution modality and mean | — |
| Non-coding RNA homology annotation | H. duospinosa genome | none | ncRNA family/tRNA isotype copy numbers | tRNAscan; Rfam |
- – Genome assembled to 375 Mbp with scaffold N50 of ~230 Kbp and ~180-225x coverage; flow cytometry estimate 320 Mbp 375 Mbp; N50 226,503 bp; max 2.8 Mbp
- – BUSCO recovery of complete genes indicates high assembly quality 95.3% complete
- – H. duospinosa transcripts show unimodal CpG[o/e] distribution like Drosophila, unlike bimodal Apis, indicating absence of historical DNA methylation mean CpG[o/e] = 0.7
- – No Dnmt1 or Dnmt3 orthologs identified; three Dnmt2 (TRDMT1) and one Tet1 ortholog present
- – 15 ionotropic glutamate and 18 gustatory receptors identified, but no odorant receptors or Orco 15 iGluRs; 18 GRs
- – 59 bacterial and 96 fungal HGT candidate genes detected on scaffolds otherwise containing arthropod genes 59 bacterial; 96 fungal
- – 23 chitinase-like genes identified in the assembly 23 genes
- – SSU rRNA contains a 1454 bp insertion (position 496-1949) encoding a reverse transcriptase homolog; eight paralogues detected 1454 bp; 8 paralogues
- other 375 Mbp genome size (Total assembly size vs 320 Mbp flow cytometry estimate)
- count N50 = 226,503 bp (Scaffold N50 of genome assembly)
- other 95.3% complete BUSCO genes (Assembly completeness (arthropoda_odb9))
- count 9911 supported gene models (Confident gene models of 12,000 total)
- other 1.56 × 10^-3 heterozygosity (Genome-wide heterozygosity)
- other 42.96% genome masked as repeats (Repeat content; DNA elements 8.42%, LTR 2.78%, class I TEs 4.37%)
- count 548 serine tRNA copies (tRNA isotype copy number vs average 18 for others)
- count 370,877 amino acids (Phylogenomic dataset placing H. duospinosa in Neanuridae)
Statistical methods review
Model: opusA neutral, descriptive read of the statistical approach — what was done, and (for shared learning, not as criticism) what could also have been done.
This is a descriptive draft-genome and transcriptome study that reports assembly metrics, gene/repeat annotations, and comparative genomic observations rather than formal hypothesis testing. Inferences are drawn from sequence-similarity searches (BLAST/TBLASTN/BLASTP with stated e-value thresholds), homology-based annotation pipelines (e.g., BUSCO, Rfam, tRNAscan), and the qualitative shape (unimodal vs. bimodal) of normalised CpG[o/e] distributions, with phylogenetic placement inferred from a concatenated amino-acid alignment. Results are presented mainly as counts, percentages, and distributional histograms; no inferential significance tests, p-values, or confidence intervals are reported.
| Test | Applied to | n | Assumptions |
|---|---|---|---|
| BLAST-based sequence similarity searches (TBLASTN, BLASTP) using e-value cut-offs | microbe/HGT screening (~0.2% TBLASTN matches), protein annotation against NCBI nr (BLASTP e-value 1e-6), repeat-model filtering (e-value 1e-5) | — | na |
| Distributional (qualitative) assessment of normalised CpG content, CpG[o/e] | Fig. 3: comparison of unimodal vs. bimodal CpG[o/e] histograms across Apis mellifera, Drosophila melanogaster, and H. duospinosa to infer historical DNA methylation | per-gene/per-transcript CpG[o/e] values and 1 kb genomic fragments (counts not fully specified) | not stated |
| Phylogenomic tree inference from a concatenated amino-acid supermatrix | placement of H. duospinosa within Neanuridae (370,877 amino acids) | 370,877 amino acid positions | not stated |
| Homology/ortholog detection and completeness assessment (BUSCO, Rfam, tRNAscan) | gene-model completeness (BUSCO arthropoda_odb9), ncRNA and tRNA identification, sex-determination and Dnmt ortholog presence/absence | 1066 conserved BUSCO genes | na |
-
Historical DNA methylation was inferred from the qualitative shape (unimodal vs. bimodal) of the CpG[o/e] distribution.↳ Could also: A formal mixture-model fit (e.g., fitting one- vs. two-component Gaussian mixtures and comparing them by likelihood ratio or BIC, or a dip test of unimodality such as Hartigan's). — A model-based comparison would attach a quantitative criterion to the unimodal-vs-bimodal call, complementing the visual assessment of the histograms.
-
Phylogenetic placement was reported from a concatenated supermatrix of 370,877 amino acids.↳ Could also: Reporting node support (e.g., bootstrap or posterior probabilities) and/or a coalescent-based species-tree analysis alongside the concatenated tree. — Support values and an alternative tree-estimation framework would convey how strongly the data favour the inferred placement and its robustness to method choice.
-
Homology and HGT/ortholog calls relied on fixed BLAST e-value thresholds.↳ Could also: Phylogeny-aware HGT detection (e.g., gene-tree/species-tree reconciliation) or profile-HMM searches with explicit score thresholds. — These approaches add an explicit evolutionary or probabilistic model to candidate calls and can help distinguish transfer from contamination or shared ancestry.
-
Summary quantities such as mean CpG[o/e] and repeat/annotation percentages were reported as point values.↳ Could also: Accompanying each summary with a measure of spread (SD, IQR) or a 95% confidence interval. — Reporting dispersion alongside means conveys the variability behind a single summary number and is often preferred for distributional quantities.
-
Comparative repeat and genome content were contrasted descriptively against reference species (e.g., D. melanogaster).↳ Could also: A phylogenetically informed comparative framework or explicit enrichment tests when comparing counts across genomes. — Such methods would account for relatedness among compared taxa and provide a quantitative basis for cross-genome differences.
Result convergence & founder nodes
Findings this paper shares with others that ran a comparable experiment. A node’s strength is how many independent papers report it (replication breadth) — not how often it is cited, so a heavily-replicated but under-cited founder still stands out.
-
Holacanthella duospinosa transcripts show a unimodal CpG[o/e] distribution (mean ~0.7), indicating absence of historical DNA methylation, contrasting with the bimodal distribution in Apis melliferaother holacanthella-duospinosa none 2017×1papers★ This paper is the founder (earliest)
-
Holacanthella duospinosa encodes 23 chitinase-like genes in its genomeWGS holacanthella-duospinosa 2017×1papers★ This paper is the founder (earliest)
-
Holacanthella duospinosa lacks DNMT1 and DNMT3 orthologs but encodes three TRDMT1 (DNMT2) paralogs and one TET1 orthologWGS holacanthella-duospinosa none 2017×1papers★ This paper is the founder (earliest)
-
Holacanthella duospinosa genome assembles to 375 Mbp with scaffold N50 of 226,503 bp, consistent with flow cytometry estimate of 320 MbpWGS holacanthella-duospinosa 2017×1papers★ This paper is the founder (earliest)
-
Holacanthella duospinosa genome contains 59 bacterial and 96 fungal horizontal gene transfer candidate genes flanked by arthropod genes on the same scaffoldsWGS holacanthella-duospinosa 2017×1papers★ This paper is the founder (earliest)
-
Holacanthella duospinosa encodes no odorant receptors or Orco co-receptor, but possesses 15 ionotropic glutamate receptors and 18 gustatory receptorsWGS holacanthella-duospinosa none 2017×1papers★ This paper is the founder (earliest)
-
Holacanthella duospinosa SSU rRNA harbors a 1454 bp reverse transcriptase insertion at position 496-1949, with eight genomic paraloguesWGS holacanthella-duospinosa 2017×1papers★ This paper is the founder (earliest)
Citation network
Where this publication sits in the reproducibility-weighted citation graph — what it is built on, and what is built on it. Citation data from OpenAlex.
Assessed papers, coloured by verdict. Click a node to open it.
- No assessed neighbours yet — the network grows as more papers are assessed.
Data lineage
The datasets this paper uses (text-mined from the full text via Europe PMC), and which other assessed papers stand on the same data. A shared dataset is a factual link — not a judgement.
What was reproduced
The exact results taken into scope, with each reported value next to the value our attempt produced.
Scope — pmid-29041914
Paper: Wu et al. 2017, Analysis of the genome of the New Zealand giant collembolan (Holacanthella duospinosa) sheds light on hexapod evolution. BMC Genomics 18:795. DOI 10.1186/s12864-017-4197-1.
Data (public): SRA BioProject PRJNA384703 (raw Illumina reads); the
assembled genome is deposited at GenBank as GCA_002738285.1 (ASM273828v1) —
this is the primary computational output of the assembly pipeline and is what we
reproduce against.
Code: https://github.com/sujaikumar/assemblage (Sujai Kumar's generic
assembly/contamination toolkit, pinned 91ddcad). This is a third-party toolkit
(blobtools/assembly helper scripts), not a turnkey author pipeline. Per brief
rule P16, applying standard third-party assembly-QC tools to the paper's own
deposited data is an equally valid reproduction. We therefore reproduce the
reported assembly-derived QC numbers by recomputing them independently with
standard tools (seqkit / assembly-stats / BUSCO) on the paper's deposited
assembly.
In scope (pipeline-derived, attempted)
| Result | Reported (Table 2 / text) | Pipeline | How we reproduce |
|---|---|---|---|
| Assembly total size | 375 Mbp text; 370,315,149 bp (>100 bp, 410,937 seqs) Table 2 | SOAPdenovo2+GapCloser+SSPACE | seqkit/assembly-stats on GCA_002738285.1 |
| Scaffold N50 | 226,503 bp (L50=317) ; "~230 Kbp" text | " | assembly-stats |
| Longest scaffold | 2,807,427 bp (Table 2) | " | assembly-stats |
| GC content | 33.40% (Table 2) | " | seqkit fx2tab / assembly-stats |
| N (gap) content | 2.18% (Table 2) | " | seqkit stats -a |
| BUSCO completeness | 825/1066 (77.4%) complete, 69 (6.5%) fragmented, arthropoda_odb9 | BUSCO v2.0.1 | BUSCO v5, arthropoda_odb10, genome mode |
Important note up front: the public deposit GCA_002738285.1 is a
filtered version of the Table-2 assembly — NCBI's submission screen drops
sub-200 bp scaffolds and flagged contaminants. NCBI's own summary reports
62,430 scaffolds / 327.6 Mbp / N50 310,226 vs the paper's 410,937 seqs /
370.3 Mbp / N50 226,503. So total size, sequence count and N50 are expected to
partially match (lower count, higher N50 after filtering); GC%, N% and the
longest scaffold should match closely. We compare against both the paper and
NCBI's reported stats and report the gap honestly (filtering, not fabrication).
Out of scope / not attempted (hard 20% or non-pipeline)
- Genome size 320 Mbp — flow cytometry (wet-lab), not a pipeline result.
- Repeat content 42.96% (Table 4) — needs de-novo RepeatModeler + RepeatMasker (many CPU-hours); deferred as the hard 20%.
- Gene annotation (12,000 / 9,911 models) — full MAKER2+Augustus+FGENESH pipeline requiring RNA-seq evidence; heavy, deferred.
- Transcriptome assembly (Trinity) — heavy, deferred.
- HGT (59 bacterial / 96 fungal genes) — manual/curated downstream analysis.
- Heterozygosity 1.56e-3 — derived during assembly QC; not cleanly re-derivable from the deposit alone.
Assessments & scoring basis
Each contributor’s verdict, the per-question basis, and the auditable, itemised worksheet behind it.
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.
Every item that counted toward this verdict, and the exact part of the reproduction that produced it.
This is a solid, explainable reproduction of a genome-assembly QC paper. The robust metrics reproduce cleanly — longest scaffold is bit-identical (2,807,427 bp), GC 33.35% vs 33.40%, BUSCO completeness 95.4% vs 95.3% — and the only large deviations (total size 327.6 vs 370.3 Mbp, N50 310,226 vs 226,503 bp) are an input-side artefact of NCBI deposit filtering (sub-200bp + contaminant removal), confirmed by exact agreement with NCBI's own published stats. The cause sits on the data-availability/preprocessing side and is technical/expected, not an authors' defect, and there is no fabrication signal. Severity is moderate in raw magnitude but the central conclusion (a high-completeness springtail genome) holds, so overall yellow rather than green only because of the explained size/N50 gaps and BUSCO tool-version drift.
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-4-8Measured resources invested to assess this paper — sanitised (machine class only, no job ids/paths). Compute = HPC accounting (SLURM); tokens = the AI agent's session.