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

Reproducible, portable, and efficient ancient genome reconstruction with nf-core/eager.

PeerJ · 2021
L1 95/100 3/4
Why this verdict

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

Reproduced on the brainbox compute brainarbeit.com
✓ What held up
  • Same input data as the authors
  • Reported values were directly comparable
  • No authors-side cause for any deviation
  • Any deviation was negligible
What did not (or only partly)
  • 🟡A deviation arose in the data or preprocessing
  • 🟡Reported values were not (fully) derivable from the shared data
  • 🟡The central claim did not (fully) hold under reproduction
  • 🟡Overall, the reproduction showed a material discrepancy
How its reproducibility compares
95/100
Reproducibility score
1.2 SD above mean
vs. all fields · 1173 studies
🎯 Scores higher than 89% of all assessed papers rank 105 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

REPRODUCED (1:1). Software/pipeline paper (nf-core/eager). Re-ran the exact documented pipeline (nf-core/eager v2.2.0dev @ commit 830c22d, Nextflow 22.10.6, conda profile with tool versions identical to the paper's singularity image: bwa 0.7.17, AdapterRemoval 2.3.1, Picard 2.22.9, samtools 1.9, DamageProfiler 0.4.9, QualiMap 2.2.2d, MultiQC 1.9) on the Viking-age cod benchmark (ENA PRJEB20524: COD076=ERR1943600-02, COD092=ERR1943607-09) mapped to gadMor3.0 GCF_902167405.1, on «our HPC» via SLURM («job» COMPLETED rc=0, 53m) with data on «infra». All 18 Table 3 / supplement mapping metrics graded: 12 EXACT (processed reads, % endogenous, duplication, mean read length, coverage, 5' C>T damage — all to 4+ sig figs), 6 within-tolerance (bwa-aln stochastic read counts, all within the <=~2 sd run-to-run variance the paper itself documents across its 10 replicate runs), 0 mismatch. Reported ground-truth recovered from the paper supplement (apeltzer/eager2-paper, 10-replicate MultiQC CSVs). Exact match of all deterministic values is strong evidence the table is genuine pipeline output, no fabrication indicators. Every reproduced value traces to reproduction/outputs/*.txt. NOT attempted (out of scope): Table 2 runtimes (hardware-specific), EAGER v1/PALEOMIX comparison, Barquera human-aDNA metagenomic demonstration. All grades provisional (must_be_independently_checked=true).

💻 Code ↗ 🗄 Data: 10.5281/zenodo.4382153

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 ⛓ 5c05b8025f49
✎ 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 paper does not test a scientific hypothesis but presents a redesigned bioinformatics tool, asking whether a completely re-implemented, Nextflow-based version of the EAGER pipeline can improve accessibility/adaptability, reproducibility, and adherence to current best practices for ancient genomic data analysis compared to existing pipelines (EAGER, PALEOMIX).

Core claims
  • nf-core/eager is a complete redesign and extension of the EAGER pipeline in Nextflow, built within the nf-core framework to ensure high-quality, sustainable software development. resource
  • nf-core/eager supports automated processing and merging of complex multi-library, multi-lane, multi-sequencing-run data via an optional TSV input table, unlike the original EAGER which handled only a single library at a time. method
  • nf-core/eager adds ancient metagenomic screening functionality (MALT, Kraken2, HOPS/MaltExtract, BBduk) for pathogen and microbiome detection in off-target reads, extending beyond the original EAGER pipeline. method
  • nf-core/eager aggregates outputs from every supported tool into a single comprehensive MultiQC report, replacing the need to inspect many independent output files as required in EAGER/PALEOMIX. method
  • nf-core/eager provides native HPC scheduler and cloud computing integration plus software container/environment support (Conda, Docker, Singularity), features absent or partial in EAGER and PALEOMIX. finding
  • New analytical modules added versus original EAGER include fastp (poly-G artefact removal), FreeBayes and pileupCaller (variant/genotype calling), Sex.DetERRmine (biological sex determination), and ANGSD (nuclear contamination estimation). method
  • A 'FASTQ host removal' script allows safe public sharing of metagenomic sequencing data by stripping reads mapped to a (human) reference, protecting personal data privacy. method
  • mapDamage2 and Schmutzi were not carried over from EAGER because DamageProfiler is a faster successor and a stable Schmutzi release was unavailable on Bioconda at time of writing. method
Experimental setups
Assay System Perturbation Readout Platform
Whole-genome sequencing / host genomic analysis and metagenomic pathogen screening (re-analysis, functionality demonstration) Human skeletal remains (three 16th century individuals from a mass burial site, Mexico City) none (re-analysis of previously published data from Barquera et al., 2020) Genetic/genomic results and putative pathogen detection
Key statistics
  • count more than 1 billion reads (Sequencing reads required for 1X depth coverage of a Yersinia pestis genome due to low endogenous DNA content, as reported in Rasmussen et al., 2015)

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 is a bioinformatics software paper describing nf-core/eager, a Nextflow-based pipeline for ancient genomic data processing. The paper does not report inferential statistics or hypothesis tests; instead it presents a qualitative feature-comparison table across three pipelines and a benchmarking/demonstration re-analysis of a previously published dataset. Quantitative outputs are computational performance and pipeline quality-control metrics rather than biological effect estimates.

Replicationunclear GroupsThree pipelines compared qualitatively (EAGER, PALEOMIX, nf-core/eager); one re-analysis of published ancient-DNA data (Barquera et al. 2020, n not stated in available text) Pairingna Randomization/blindingna Dispersionnone
Approaches that could also have been used
  • Pipeline feature comparison across EAGER, PALEOMIX, and nf-core/eager was reported as a binary Yes/No/Partial table (Table 1) without quantitative performance metrics in the feature-comparison section
    Could also: A structured benchmark with quantitative runtime, memory consumption, and throughput metrics (e.g., wall-clock time per sample, peak RAM) across the three pipelines on identical test datasets could also be reported — Quantitative benchmarks allow readers to estimate resource costs on their own hardware and make empirically grounded tool choices; binary feature tables convey capability breadth but not efficiency trade-offs
  • Pipeline reproducibility is demonstrated by re-running a published dataset (Barquera et al. 2020) and implicitly comparing outputs, but no formal concordance or agreement statistics are reported in the available text
    Could also: Concordance metrics such as genotype concordance rates, Pearson or Spearman correlation of coverage estimates, or Bland–Altman analysis between pipeline outputs could also be reported — Formal agreement statistics quantify the degree to which the new pipeline reproduces prior results, providing a concrete reproducibility claim rather than a qualitative assertion
  • Software dependency management relies on Conda, Docker, and Singularity containers, with tool versions drawn from Bioconda at time of release; specific tool versions are not tabulated in the paper body
    Could also: A versioned dependency table or a machine-readable software bill of materials (e.g., SBOM or Conda lock file excerpt) could also be included as a supplementary resource — Pinned version tables allow exact replication of the computational environment years after publication, which is particularly important for pipelines where upstream tool behaviour can change between releases
  • Ancient DNA damage and authenticity metrics (e.g., C-to-T substitution rates from DamageProfiler) are reported as pipeline outputs for the demonstration dataset but their uncertainty is not quantified in the available text
    Could also: Bootstrap confidence intervals or Bayesian credible intervals on damage-rate estimates could also be reported, as implemented in tools such as mapDamage2's Bayesian mode — Interval estimates on damage rates allow users to distinguish genuine aDNA signal from sampling noise, particularly for low-coverage libraries where point estimates are highly variable
  • Metagenomic profiling results from MALT and Kraken2 are presented as pipeline outputs without a cross-tool concordance assessment in the available text
    Could also: A sensitivity/specificity or F1-score comparison between the two profilers on a mock community or known-composition sample could also be reported — Tool-concordance data on a reference community helps users select or combine profilers appropriately for their target organisms and coverage levels
Software: Nextflow · FastQC · AdapterRemoval2 · fastp · BWA (aln / mem) · Bowtie2 · SAMtools · Picard MarkDuplicates · DeDup · PreSeq · Qualimap2 · bedtools · Sex.DetERRmine · DamageProfiler · PMDtools · mapDamage2 · ANGSD · GATK (UnifiedGenotyper / HaplotypeCaller) · FreeBayes · MultiVCFAnalyzer · MALT · Kraken2 · MaltExtract / HOPS · MultiQC · Conda / Docker / Singularity

What was reproduced

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

Scope — pmid-33777521 (nf-core/eager paper, Fellows Yates et al. 2021, PeerJ)

Paper: "Reproducible, portable, and efficient ancient genome reconstruction with nf-core/eager." PMID 33777521 · PMC7977378 · DOI 10.7717/peerj.10947.

This is a software/pipeline paper. The code artifact is the Nextflow pipeline nf-core/eager (NOT stschiff/sequenceTools, which is only one bundled tool). Benchmarking version pinned in the paper: v2.2.0dev, commit 830c22d.

In-scope (pipeline-derived, reproducible)

Table 3 — mapping statistics for the Viking-age cod (Gadus morhua) benchmark, produced by running nf-core/eager on two samples (3 Illumina PE runs each) from Star et al. 2017 (ENA PRJEB20524), mapped to gadMor3.0 (GCF_902167405.1). Reproduced by re-running the exact documented pipeline command:

nextflow run nf-core/eager -r dev --input <tsv> -c pub_eager_vikingfish.conf \
  -profile pub_eager_vikingfish,singularity \
  --fasta GCF_902167405.1_gadMor3.0_genomic.fasta \
  --skip_fastqc --skip_preseq --run_bam_filtering \
  --bam_mapping_quality_threshold 25 --bam_discard_unmapped \
  --bam_unmapped_type 'discard' --dedupper 'markduplicates'

All other params = eager v2.2.0 defaults (mapper bwaaln, n=0.04 k=2 l=1024; AdapterRemoval collapse, clip_readlength 30, q20; Picard MarkDuplicates; DamageProfiler; QualiMap; endorSpy). The paper ran this 10× to get run-to-run variance; the per-replicate MultiQC general-stats CSVs are in the paper supplement (github.com/apeltzer/eager2-paper, content/supplement/benchmarking/results/) and are used here as the ground-truth reported values (see original/reported_values_table3.tsv).

Target metrics (per sample COD076, COD092): processed reads (post-AdapterRemoval), mapped reads (pre-filter), MAPQ>=25 filtered mapped reads, % endogenous DNA, % duplication, mean read length, post-dedup mapped reads, mean coverage, 5' C>T damage.

Input data (ENA PRJEB20524, Gadus morhua, HiSeq 2500, PAIRED)

  • COD076 (lib COD076E1bL1): ERR1943600 (L8), ERR1943601 (L6), ERR1943602 (L1)
  • COD092 (lib COD092E1bL1i69): ERR1943607 (L6), ERR1943608 (L7), ERR1943609 (L8)
  • Reference: GCF_902167405.1_gadMor3.0_genomic.fna (RefSeq, NCBI FTP)

Out of scope (not pipeline-deterministic / not attempted)

  • Table 2 runtimes (105.6 / 209.2 min etc.) — hardware-specific (GWDG cloud, 32 cores), NOT reproducible on different hardware. Recorded but not graded.
  • EAGER v1 and PALEOMIX comparison numbers — require those other pipelines; tangential to reproducing nf-core/eager's own output.
  • Barquera et al. 2020 human-aDNA demonstration (sex determination, pathogen screening with the full Oct-2017 NCBI nt database, zenodo 4382153) — extremely heavy (human genome mapping + metagenomic screen against full nt); deferred / likely not attempted. Qualitative claims (all male; T. pallidum + HBV detected) are not single reproducible numbers.

Execution

All compute on «our HPC» («infra») via SLURM, data on «infra». nf-core/eager run with the pinned conda environment.yml (same pinned tool versions as the singularity image: bwa 0.7.17, AdapterRemoval 2.3.1, picard 2.22.9, samtools 1.9, damageprofiler 0.4.9, qualimap 2.2.2d, multiqc 1.9). Singularity unavailable on «our HPC» without root → conda profile with identical pinned versions.

Figures / tables: Table
T3_COD076_processed
Reported
71388991
Reproduced
71388991
exact
T3_COD076_mapped_prefilter
Reported
25136796
Reproduced
25136558
within tolerance
T3_COD076_mapped_filt
Reported
16786514.6
Reproduced
16786396
within tolerance
T3_COD076_endogenous
Reported
23.5142
Reproduced
23.513984
exact
T3_COD076_duplication
Reported
0.2787
Reproduced
0.278744
exact
T3_COD076_readlen
Reported
49.4422
Reproduced
49.44219
exact
T3_COD076_dedup_mapped
Reported
12107310.1
Reproduced
12107287
within tolerance
T3_COD076_coverage
Reported
0.8935
Reproduced
0.8935
exact
T3_COD076_damage5p
Reported
0.1437
Reproduced
0.143687
exact
T3_COD092_processed
Reported
69615709
Reproduced
69615709
exact
T3_COD092_mapped_prefilter
Reported
25207381.3
Reproduced
25207434
within tolerance
T3_COD092_mapped_filt
Reported
16283194.7
Reproduced
16283269
within tolerance
T3_COD092_endogenous
Reported
23.3901
Reproduced
23.390222
exact
T3_COD092_duplication
Reported
0.1605
Reproduced
0.160525
exact
T3_COD092_readlen
Reported
48.7593
Reproduced
48.75912
exact
T3_COD092_dedup_mapped
Reported
13669322.7
Reproduced
13669394
within tolerance
T3_COD092_coverage
Reported
0.9948
Reproduced
0.9948
exact
T3_COD092_damage5p
Reported
0.0593
Reproduced
0.059278
exact

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 95/100

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)
🤝
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.

355.1 k
tokens (I/O) · 31.4 M incl. cache
179 min
runtime
Per-job HPC accounting not captured for this run — the runtime shown is the reproduction’s measured wall-clock time.