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

Starvation-induced transgenerational inheritance of small RNAs in C. elegans.

Cell · 2014
L1 64/100 3/4
Why this verdict

The main results reproduced, with only marginal, non-material deviations.

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: 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 🟡
Total score +4
✓ What held up
  • Same input data as the authors
  • Reported values were directly comparable
  • Reported values are derivable from the shared data
  • Any deviation was negligible
  • The central claim held under reproduction
What did not (or only partly)
  • 🟡A deviation arose in the data or preprocessing
  • 🟡A deviation was attributed to the published material
  • 🟡Overall, the reproduction showed a material discrepancy
How its reproducibility compares
64/100
Reproducibility score
0.6 SD below mean
vs. all fields · 1173 studies
🎯 Scores higher than 25% of all assessed papers rank 854 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 the core pipeline-derived findings of Rechavi et al. 2014 (PMID 25018105) using an independently-written bowtie2+htseq-count+DESeq2 pipeline on the paper's own SRA data (PRJNA259320, 36/36 runs downloaded and MD5-verified). The paper's central transgenerational-inheritance claim (F3-of-starved retains a starvation-like 22G-RNA signature) reproduces well: down-direction inheritance 54.3% observed vs 52.0% reported, up-direction 31.1% vs 26.3%. The underlying P0 fed-vs-starved 22G-RNA differential expression is within ~15% of the paper's 578up/597down (655/693 observed); the 26G-RNA result is a partial match (down-direction close, up-direction off ~10x on a small gene set). mRNA-seq and hrde-1/rde-4-mutant P0-generation comparisons were computed as supporting/exploratory data but are not direct numeric reproductions of a specific paper-reported figure (the paper's claims there are qualitative or concern F3-generation transmission, which was not fully re-tested for the mutants). GO enrichment (GOrilla) and lifespan phenotype assays are explicitly out of scope. Three non-obvious pipeline bugs (a POSIX $ENV variable collision, a wrong small-RNA 3' adapter, and an htseq-count '-t gene' feature-type mismatch against a GTF vintage with no 'gene' lines) were each caught by inspecting real output content/size rather than trusting SLURM COMPLETED status, and fixed before the final results reported here.

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.

✎ 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
2026-07-29
Rubric version
v1.0
Assessed by
🤖 AI curator · claude (ai-curator room) · v1.0 · run #1 2026-07-31
no human curator yet
Last updated
2026-07-31

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: opus
Founding hypothesis

The authors ask whether a natural environmental change—starvation-induced L1 developmental arrest—can elicit an endogenous small RNA response that is transgenerationally inherited in C. elegans, rather than only the previously described heritable silencing directed against foreign DNA.

Core claims
  • L1 starvation induces changes in endogenous 22G small RNAs (STGs) that are inherited for at least three generations in fed descendants. finding
  • This is the first demonstration that an endogenous, environmentally induced response (not foreign DNA) triggers heritable small RNA inheritance targeting endogenous genes. finding
  • The inherited starvation-induced small RNAs preferentially target genes with nutrition-related functions, most significantly nutrient reservoir activity, including all six vitellogenins and 20 F-box proteins. finding
  • Transgenerational transmission of the starvation-induced small RNAs requires the germline nuclear argonaute HRDE-1. mechanism
  • The heritable small RNA response is largely dependent on the dsRNA-binding DICER cofactor RDE-4. mechanism
  • F3 progeny of animals starved as L1s show a significantly increased lifespan, corroborating a transgenerational memory of dietary history. finding
  • mRNA targets of inherited HRDE-1-bound small RNAs are downregulated while targets of inherited CSR-1-bound small RNAs are upregulated, consistent with opposing silencing/licensing roles. mechanism
  • Definition and use of 'STGs' (clusters of small RNAs aligning antisense to a given gene) as an analytic unit, together with a cloning protocol capturing secondary 22G endo-siRNAs. method
Experimental setups
Assay System Perturbation Readout Platform
Small RNA sequencing (protocol enabling cloning of secondary/RdRP-amplified endo-siRNAs) C. elegans wild-type (N2) young adults 6-day L1 starvation versus continuous feeding 22G and 26G small RNA abundance mapped antisense to genes (STGs); differential expression by DESeq2, FDR < 0.1
Small RNA sequencing C. elegans N2 F3 progeny (fed for three generations) of P0s starved as L1s ancestral L1 starvation (P0), descendants fed inherited differential 22G STG levels; clustering/PCA relative to fed and starved P0
Small RNA sequencing C. elegans rde-4 mutants (P0 fed/starved and F3) rde-4 loss-of-function combined with L1 starvation number of STGs induced and inherited versus N2; clustering pattern
Small RNA sequencing C. elegans hrde-1 mutant hermaphrodites (fertile at nonrestrictive temperature; P0 and F3) hrde-1 loss-of-function combined with L1 starvation retention/loss of inherited STGs; clustering of fed, starved and F3 samples
mRNA sequencing (transcriptome sequencing) C. elegans N2 young adults (fed, starved-as-L1 P0) and F3 progeny; also rde-4 animals L1 starvation / genotype mRNA levels of putative targets of differentially expressed and heritable STGs
Lifespan assay C. elegans great-grandprogeny (F3) of starved versus continuously fed animals ancestral L1 starvation survival/lifespan, three independent experiments
Computational analyses: principal components analysis, DESeq2 differential expression, GOrilla GO enrichment, hypergeometric overlap testing, cross-referencing to published HRDE-1/CSR-1-bound and germline-expressed gene sets small RNA and mRNA sequencing datasets from N2, rde-4, hrde-1 worms none (in silico) sample clustering, differentially expressed STGs, functional enrichment, overlap significance and fold enrichment
Key results
  • L1 starvation changes 22G STG levels in P0 young adults (578 upregulated, 597 downregulated), with smaller changes in 26G STGs (20 up, 38 down) 578 up / 597 down (22G); 20 up / 38 down (26G)
  • A large fraction of starvation-induced STG changes persist in fed F3 progeny: 26.3% (152/578) of upregulated and 52% (311/597) of downregulated STGs remain changed 152/578 (26.3%); 311/597 (52%)
  • Putative targets of differentially expressed and heritable STGs are most significantly enriched for nutrient reservoir activity; all six vitellogenins and 20 F-box protein genes are targeted by heritable STGs vitellogenins 6/6, 38.7-fold enrichment; F-box 5-fold enrichment
  • In hrde-1 mutants 96.8% of heritable STGs are no longer inherited (only 13/463 retained), and fed parents cluster with F3 rather than with starved parents 13/463 retained (96.8% lost)
  • In rde-4 mutants 93% of inheritance is lost (only 33/463 STGs inherited) and mRNA changes among heritable targets drop from 96/463 (20%) in N2 to 33 (7%) 33/463 (7%) versus 96/463 (20%)
  • F3 descendants of starved animals live significantly longer than descendants of continuously fed animals in three independent experiments 37%, 70% and 22% lifespan increase
  • mRNA levels of 88 genes predicted to be targets of starvation-upregulated 22G STGs were downregulated in P0 adults, while 14 predicted targets were upregulated 88 down / 14 up
  • Inherited HRDE-1-bound-small-RNA targets are mostly downregulated (17/26, 65.3%) whereas CSR-1-bound targets are mostly upregulated (11/16, 68.7%), and inherited CSR-1 STG levels correlate strongly with germline-expressed cognate mRNAs R2 = 0.907
Key statistics
  • correlation R2 = 0.907, p value < 0.00026 (inherited CSR-1-bound STG levels versus germline-expressed cognate mRNA targets)
  • pvalue p value < 1.256 × 10−292 (hypergeometric overlap of STGs downregulated in P0 starved and still downregulated in F3 (311/597))
  • pvalue p value < 1.399 × 10−71 (hypergeometric overlap of STGs upregulated in P0 starved and still upregulated in F3 (152/578))
  • fold_change 38.7-fold enrichment, p value < 2.950 × 10−10 (heritable STGs aligning to all six vitellogenin genes)
  • fold_change 26.1-fold enrichment, p value < 5.629 × 10−29 (91% of 22G STGs overlapping 26G STGs transgenerationally downregulated in F3)
  • pvalue FDR < 8.61 × 10−4 (GOrilla enrichment for nutrient reservoir activity among targets of differentially expressed 22G STGs in P0)
  • pvalue 37% increase p < 0.0001; 70% increase p < 0.0001; 22% increase p = 0.004 (lifespan increase of F3 descendants of starved animals, three independent experiments)
  • fold_change 3.7-fold enrichment, p value < 1.139 × 10−40 (25.9% of inherited STGs previously shown to physically bind HRDE-1)

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.

The study used high-throughput small RNA and mRNA sequencing across parental (P0) and F3-generation C. elegans, comparing fed versus starved conditions and several mutant backgrounds (rde-4, hrde-1). Differential expression of small RNA species was assessed with DESeq2 at an FDR threshold, group separation was visualized with PCA, and enrichment of gene sets/overlaps between differentially expressed gene lists was assessed with hypergeometric/enrichment tests (including GOrilla for GO terms). Lifespan differences across generations were evaluated in three independent experiments with reported p values, and results throughout are reported largely as point estimates, fold-enrichments, and p values rather than with explicit measures of dispersion.

Replicationbiological Sample sizePCA is described as showing three biological replicates clustering together per condition; lifespan assays are described as three independent experiments; specific worm counts per condition/assay are not given in the provided text Groupsfed vs. L1-starved P0 worms; F3 fed progeny of starved vs. fed ancestors; N2 vs. rde-4 vs. hrde-1 mutant backgrounds; lifespan of F3 descendants of starved vs. fed great-grandparents Pairingunpaired Randomization/blindingnot stated Dispersionunclear Exact p-valuesyes Effect sizesyes Confidence intervalsno Multiplicity correctionFalse discovery rate (FDR) thresholds, e.g. DESeq2 FDR < 0.1 and GOrilla FDR < 8.61 × 10⁻⁴; specific underlying correction procedure (e.g., Benjamini-Hochberg) not explicitly named in the text
Statistical tests used
Test Applied to n Assumptions
Principal components analysis (PCA) clustering of small RNA pools from N2, rde-4, and hrde-1 samples across conditions (Figure 1A, Figure S1, Figure S5) three biological replicates per condition, as stated not stated
DESeq2 (differential expression, FDR-based) comparison of 22G and 26G small RNA (STG) levels between fed and starved P0 animals (Figures 2A, S2) three biological replicates (per PCA description) not stated
Hypergeometric test overlap between P0-differential STGs and F3-differential STGs (Figures 3A, 3B; Table S2), and overlap with HRDE-1-bound small RNA sets gene list sizes as given (e.g., 578, 597, 463 STGs) not stated
GO term enrichment analysis (GOrilla, FDR-based) functional enrichment of putative STG target genes (Figures 2B, S3, 3C) gene lists of putative targets (e.g., 1,175; 463 genes) not stated
Fold-enrichment/depletion test (hypergeometric or Fisher's-exact-type, exact method not named) vitellogenin family enrichment, F-box protein enrichment, germline-expression enrichment among STG targets (Figure 3B, 3C) gene set sizes as given (e.g., 6/6, 20 genes, 463 targets) not stated
Correlation analysis (R² reported) relationship between inherited CSR-1 STG levels and germline-expressed cognate mRNA targets (Figure 4A) not stated not stated
Approaches that could also have been used
  • Differential small RNA expression between fed and starved conditions was assessed with DESeq2 using an FDR < 0.1 threshold.
    Could also: edgeR or limma-voom, which use related but distinct dispersion-estimation and testing frameworks for count data — Comparing results across tools with different dispersion models can help characterize how sensitive the differential-expression calls are to the specific statistical model used, which is a routine cross-check in RNA-seq-type analyses.
  • Overlap significance between gene lists (e.g., P0 vs. F3 differentially expressed STGs, or STGs overlapping known binding sets) was assessed with hypergeometric tests.
    Could also: Fisher's exact test, which is mathematically closely related and commonly used interchangeably for 2x2 contingency-style overlap analyses — Fisher's exact test is a standard alternative for the same type of overlap question and can be a useful cross-check, particularly when list sizes are small.
  • Functional enrichment of target gene sets was evaluated with GOrilla using ranked/threshold-based lists.
    Could also: Gene Set Enrichment Analysis (GSEA), which uses the full ranked gene list rather than a hard significance cutoff — A rank-based approach avoids the need to choose a specific inclusion threshold and can capture enrichment signal distributed across many moderately changed genes.
  • The relationship between inherited CSR-1 small RNA levels and cognate mRNA levels was summarized with an R² value and a single p value.
    Could also: Reporting the correlation coefficient (Pearson's r or Spearman's rho) together with its 95% confidence interval — A confidence interval around the correlation coefficient conveys the precision of the estimate in addition to statistical significance, which can be informative for smaller sample sizes.
  • Lifespan differences across generations were assessed in three independent experiments with associated p values, though the specific test was not named in the excerpt.
    Could also: The log-rank (Mantel-Cox) test, a standard method for comparing full survival curves, potentially supplemented with a Cox proportional hazards model — Log-rank/Cox approaches use the entire survival distribution rather than a single summary statistic and can also yield a hazard ratio with a confidence interval to express the magnitude of the lifespan effect.
  • Results throughout are reported as point estimates, fold-changes, and p values without an explicit measure of variability such as SD, SEM, or confidence intervals.
    Could also: Including SD or 95% confidence intervals alongside point estimates — Explicit dispersion measures let readers gauge the precision of estimates independent of p value magnitude, which is often reported alongside significance testing in similar transgenerational and expression studies.
Software: DESeq2 · GOrilla

What was reproduced

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

srna_22G_P0_fed_vs_starved
Reported
{'up': 578, 'down': 597, 'source': 'paper main text'}
Reproduced
{'up': 655, 'down': 693, 'source': 'out/srna_results/summary.tsv «job»'}
within tolerance
srna_26G_P0_fed_vs_starved
Reported
{'up': 20, 'down': 38, 'source': 'paper main text'}
Reproduced
{'up': 2, 'down': 29, 'source': 'out/srna_results/summary.tsv «job»'}
partial
srna_22G_F3_inheritance_overlap
Reported
{'pct_up_inherited': 26.3, 'pct_down_inherited': 52.0, 'source': 'paper main text (152/578 up, 311/597 down)'}
Reproduced
{'pct_up_inherited': 31.1, 'pct_down_inherited': 54.3, 'source': 'out/srna_results/directional_overlap.txt «job» (204/655 up, 376/693 down)'}
within tolerance
srna_22G_hrde1_rde4_P0_fed_vs_starved
Reported
{'note': 'paper reports qualitative loss/attenuation of F3 transmission for these mutants, no specific P0-generation count to match'}
Reproduced
{'hrde-1_up': 2394, 'hrde-1_down': 2514, 'rde-4_up': 414, 'rde-4_down': 793, 'source': 'out/srna_results/summary.tsv «job»'}
partial
mrna_seq_expression_changes
Reported
{'note': 'paper reports qualitative GO/nutrition-gene enrichment and CSR-1/HRDE-1 target trends, not a specific total up/down gene count'}
Reproduced
{'N2_P0fed_vs_starved': {'up': 935, 'down': 2615}, 'N2_F3_vs_P0fed': {'up': 138, 'down': 5230}, 'hrde-1_P0fed_vs_starved': {'up': 246, 'down': 782}, 'hrde-1_F3_vs_P0fed': {'up': 809, 'down': 3364}, 'rde-4_P0fed_vs_starved': {'up': 142, 'down': 799}, 'rde-4_F3_vs_P0fed': {'up': 564, 'down': 820}, 'source': 'out/mrna_results/summary.tsv «job»'}
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 64/100

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: 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 🟡
Total score +4

This is a good, honest partial reproduction. Using the authors' own public SRA data (PRJNA259320, 36/36 runs, MD5-exact) and an independently written pipeline, the paper's headline transgenerational claim reproduces closely: direction-specific F3 22G-RNA inheritance of 54.3% down (paper 52.0%) and 31.1% up (paper 26.3%), on P0 DE sets of 655 up / 693 down vs the reported 578/597 (~1.13-1.16x). The deviations sit on our side — self-chosen adapter, 22G/26G length+5'-G classification windows, WS220 GTF and -t exon counting — not on the authors' side, and there is no fabrication signature. It falls short of green overall because the 26G up-count misses ~10x (2 vs 20, small noisy set) and because material parts of the paper (hrde-1/rde-4 F3 inheritance failure, GOrilla enrichment, lifespan assays) were never tested.

🤝
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.