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

Neuronal Small RNAs Control Behavior Transgenerationally.

Cell · 2019
L1 49/100 3/4
Why this verdict

The main result did not reproduce in this reproduction attempt. Where our recomputation produced values that differ from the published ones, those discrepancies are listed below. This is a single automated attempt — not peer review and not a finding of error or misconduct — and differences can also arise from data access, undocumented parameters or the computing environment. The verdict can be contested via “report an error”.

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: 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 🟡
Total score +9
✓ What held up
  • Same input data as the authors
What did not (or only partly)
  • 🟡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
49/100
Reproducibility score
1.4 SD below mean
vs. all fields · 1173 studies
🎯 Scores higher than 7% of all assessed papers rank 1081 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 pipeline-derived (bioinformatic) computational results of PMID 31178120 by resuming a prior disconnected room's completed download/alignment/counting work on «infra» scratch, then running a 22-contrast DESeq2 differential-expression analysis (7 mRNA + 15 small-RNA contrasts spanning genotype x tissue x generation x condition, mirroring the paper's genetic design) covering Figs 1, 2, 4 and 5 (Fig 3 wet-lab assays and Fig 6 behavioral assays are explicitly out of scope). Along the way, root-caused and fixed a genuine pipeline bug: 12/64 small-RNA libraries (a distinct HiSeq2500 sequencing batch) used a different 3' adapter (Illumina TruSeq Small RNA RA3, TGGAATTCTCGGGTGCCAAGG) than the one the original pipeline searched for, silently discarding all their reads; re-trimming with the correct adapter recovered real count data for all 12. Also fixed a DESeq2 size-factor edge case on sparse small-RNA count data via the standard poscounts fallback. After both fixes, all 22/22 contrasts completed with 0 errors. The paper's central, most specific claim -- germline downregulation of saeg-2 by neuronal RDE-4 rescue, sid-1-independently, persisting (at least partially) into the F3 generation -- is strongly and directly reproduced across multiple independent contrasts (within-tol). Broader qualitative claims (neuronal STG generation, germline STG induction from neurons, transgenerational persistence, sid-1-independence at the genome-wide level) are reproduced directionally (partial), while the paper's exact published gene-list sizes (46/124/40/18/7/5) are not independently re-derived because the paper's own STG-calling pipeline was not available to re-run -- this methodology gap is called out explicitly per claim rather than glossed over. The GSE124049 dataset (91 runs: 27 mRNA + 64 small-RNA) is complete and byte-exact vs ENA records (quality_grade B, downgraded from A only for the adapter-batch inconsistency described above, which was fully resolvable).

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

Can a neuronal process, specifically the biogenesis of RDE-4-dependent endogenous small RNAs in neurons, breach the Weismann Barrier and transmit heritable information to the germline that controls gene expression and behavior of the progeny for multiple generations?

Core claims
  • Neuron-specific synthesis of RDE-4-dependent small RNAs regulates germline amplified endogenous siRNAs and germline gene expression for multiple generations finding
  • Production of small RNAs in neurons controls the chemotaxis behavior of the progeny for at least three generations via the germline Argonaute HRDE-1 finding
  • The conserved gene saeg-2 is a target of these small RNAs and is transgenerationally downregulated in the germline; its silencing following neuronal small RNA biogenesis is required for chemotaxis under stress mechanism
  • Rescue of RDE-4 exclusively in neurons of rde-4(ne299) mutants upregulates 476 STGs bearing the hallmarks of RDE-4-dependent endo-siRNAs (DCR-1-dependent, ERGO-1-bound, somatic MUT-16-dependent) finding
  • Small RNAs can be sequenced from an isolated somatic tissue of C. elegans (FACS-sorted neurons), which had not been done before, defining a set of neuronal STGs (NeuroSTGs) method
  • The germline of Psng-1::rde-4 worms is devoid of functional RDE-4, as shown by four complementary methods, so heritable effects originate in the soma finding
  • Downregulation of 744 STGs upon neuronal RDE-4 rescue is indirect, attributed to competition between small RNA pathways over shared biogenesis factors mechanism
  • A set of 64 small RNA libraries and 27 mRNA libraries from whole worms, dissected gonads and sorted neurons is provided as a community resource (GEO: GSE124049) resource
Experimental setups
Assay System Perturbation Readout Platform
small RNA sequencing whole C. elegans adults: N2 wild type, rde-4(ne299) mutants, single-copy Psng-1::rde-4 and high-copy Psng-1/Prgef-1::rde-4 neuronal rescue strains rde-4 loss-of-function mutation with neuron-specific rescue (transgenic OE) normalized STG (small RNAs targeting a gene) read counts, RPM, differential expression Illumina-compatible small RNA libraries
mRNA sequencing (RNA-seq) whole C. elegans; N2, rde-4 mutants, Psng-1::rde-4 rescue rde-4 mutation / neuronal rde-4 rescue mRNA levels of STG-targeted genes (log2 RPM), differential expression
FACS isolation of neurons followed by small RNA and mRNA sequencing isolated Prab-3::rfp-marked neurons (single-cell suspensions) from adult N2, rde-4(-), and Psng-1::rde-4 worms rde-4 mutation / neuronal rde-4 rescue NeuroSTG expression levels and neuronal mRNA levels (log2 rpm) chemomechanical disruption protocol adapted from Kaletsky et al. (2016) followed by flow-cytometry cell sorting
mRNA deep sequencing of dissected gonads dissected gonads of Psng-1::rde-4 worms (three independent replicate gonad samples) neuronal rde-4 rescue in rde-4(ne299) background reads spanning the rde-4(ne299) insertion site (chr-III: 10,218,186) to test for functional rde-4 transcripts in the germline
single molecule fluorescent in situ hybridization (smFISH) whole worms expressing integrated single-copy Psng-1::rde-4::SL2::yfp or Prgef-1::rde-4::SL2::yfp transgenic neuronal rescue construct yfp transcript molecules in the neuronal ventral cord versus the germline; DAPI nuclei staining
fluorescence microscopy of a trans-spliced YFP reporter transgenic C. elegans expressing Psng-1::rde-4::SL2::yfp or Prgef-1::rde-4::SL2::yfp (co-injected with Punc122::GFP coelomocyte marker) transgenic neuronal rescue construct spatial pattern of YFP fluorescence (neurons versus germline)
germline RNAi feeding assay (embryonic lethality) C. elegans of indicated genotypes, including rde-4(+) controls and Psng-1::rde-4 worms feeding dsRNA-producing bacteria targeting pos-1 and mel-26, versus empty-vector control percentage of hatched eggs per plate
chemotaxis behavioral assay C. elegans progeny of neuronal RDE-4 rescue worms across generations; HRDE-1 and saeg-2 backgrounds neuronal rde-4 rescue; hrde-1 dependency; saeg-2 silencing; stress conditions chemotaxis behavior across at least three generations
Key results
  • Neuronal RDE-4 rescue upregulated 476 STGs and downregulated 744 STGs in whole worms 476 up / 744 down STGs
  • The 476 upregulated STGs are enriched for DCR-1-dependent siRNAs 5.3x enrichment
  • The 476 upregulated STGs are enriched for ERGO-1-bound endo-siRNAs 8.9x enrichment
  • The 476 upregulated STGs are enriched for endo-siRNAs that depend on somatic MUT-16 3.5x enrichment
  • 46 RDE-4-dependent NeuroSTGs were identified in sorted neurons of Psng-1::rde-4 versus rde-4(ne299); 9 of the 46 target genes also changed at the mRNA level 46 NeuroSTGs; 9/46 genes with mRNA changes
  • No functional rde-4 transcripts were detected in gonads of Psng-1::rde-4 worms: all reads spanning the ne299 insertion site carried the disabling insertion 100% (30 of 30 reads)
  • Germline RNAi against pos-1 and mel-26 was 100% efficient in rde-4(+) controls (no eggs hatched) but showed no detectable germline RNAi activity in Psng-1::rde-4 worms 100% efficiency in rde-4(+) controls
  • Neuronal mRNA data from sorted N2 neurons correlated strongly with previously published neuronal mRNA data (Kaletsky et al., 2016), and genes targeted by NeuroSTGs were enriched for neuronal GO terms rho = 0.6-0.74 across replicates; 6 of 7 top GO terms neuronal
Key statistics
  • fold_change 5.3x enrichment, p < 0.001 (Enrichment of upregulated STGs for DCR-1-dependent siRNAs after neuronal RDE-4 rescue)
  • fold_change 8.9x enrichment, p < 10^-4 (Enrichment of upregulated STGs for ERGO-1-bound endo-siRNAs)
  • fold_change 3.5x enrichment, p < 10^-4 (Enrichment of upregulated STGs for somatic MUT-16-dependent endo-siRNAs)
  • correlation rho = 0.6-0.74, p < 10^-300 (N2 sorted-neuron mRNA levels versus Kaletsky et al. (2016) neuronal mRNA data, across replicates)
  • count 476 upregulated and 744 downregulated STGs (Deseq2, adjusted p value < 0.1) (Whole-animal small RNA changes upon single-copy Psng-1::rde-4 rescue)
  • count 46 differentially expressed NeuroSTGs (Deseq2, adjusted p value < 0.1); 9 with concordant mRNA changes (FACS-sorted neurons, Psng-1::rde-4 versus rde-4(ne299))
  • count 100% (30 out of 30) reads contained the disabling ne299 insertion (Gonad mRNA reads spanning rde-4 insertion site (chr-III: 10,218,186), three replicate gonad samples)
  • count 64 small RNA libraries and 27 mRNA libraries sequenced; 412 genes targeted by NeuroSTGs with RPM > 5 (Overall sequencing resource deposited under GEO: GSE124049)

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 profiles small RNA and mRNA populations by Illumina sequencing across C. elegans genotypes (N2 wild type, rde-4 mutants, neuron-specific rde-4-rescue transgenics), using DESeq2 to call differentially expressed small-RNA-targeted genes (STGs) and mRNAs (adjusted p < 0.1). Functional enrichment of STG target lists was assessed with a permutation test (10,000 randomly drawn size-matched gene sets) and GO term enrichment via GOrilla with FDR control. A hatching-based RNAi assay comparing genotypes was analyzed with two-way ANOVA and Tukey's post hoc correction, and cross-dataset reproducibility of neuronal mRNA levels was assessed by correlation (rho). Results are shown largely as dot plots of biological replicates and bar plots of mean ± SD, with significance reported mainly via p-value thresholds/star notation.

Replicationbiological Sample sizeDescribed qualitatively as 'multiple biological replicates,' 'at least three independent experiments,' and 'n ≥ 2 biological replicates'; no formal power/sample-size calculation stated GroupsN2 wild type vs. rde-4(ne299) mutants vs. neuron-specific rde-4-rescue transgenics (Psng-1::rde-4, Prgef-1::rde-4), including isolated neurons and gonads Pairingunpaired Randomization/blindingnot stated DispersionSD Exact p-valuesno Effect sizesyes Multiplicity correctionDESeq2 adjusted p-values for STG/mRNA differential expression; Tukey's post hoc correction following two-way ANOVA; FDR threshold for GOrilla GO term enrichment
Statistical tests used
Test Applied to n Assumptions
DESeq2 (differential expression, adjusted p-value) STG and mRNA comparisons between rde-4 mutants and neuronal rde-4-rescue worms (Figures 1D, 2B, 2D, S1D, S1E) not stated
Permutation/resampling test (10,000 random size-matched gene sets) Enrichment/depletion of upregulated and downregulated STGs for DCR-1-, ERGO-1-, and MUT-16-associated gene sets (Figures 1E, 2C, S1F, S1G) not stated
Two-way ANOVA with Tukey's post hoc correction Percentage of hatched eggs following RNAi against pos-1/mel-26 across genotypes (Figure 3A; Figure S3A) at least three independent experiments per group; Figure S3A notes n ≥ 2 biological replicates not stated
Correlation analysis (rho) Comparison of measured N2 neuronal mRNA levels to previously published data (Figure S2A) four replicate samples of N2 worms not stated
GO term enrichment analysis (GOrilla) with FDR correction Genes targeted by NeuroSTGs versus whole-animal STGs (Figure S2C) not stated
Approaches that could also have been used
  • Differential expression of STGs and mRNAs was assessed with DESeq2 using an adjusted p-value cutoff of 0.1.
    Could also: edgeR or limma-voom could also be used for count-based RNA-seq differential expression analysis. — These tools use different dispersion-estimation and normalization strategies and are commonly used alongside or in place of DESeq2, which can serve as a cross-check on the called differential genes, particularly with modest replicate numbers.
  • Enrichment of STG target categories (DCR-1-, ERGO-1-, MUT-16-associated) was tested with a custom permutation approach using 10,000 randomly drawn, size-matched gene sets.
    Could also: A hypergeometric test, Fisher's exact test, or established enrichment tools (e.g., GSEA, clusterProfiler) could also be used. — These provide analytic (non-simulated) p-values and are widely used standard frameworks, which can make enrichment results directly comparable across studies using the same tools.
  • Hatching percentages across genotypes following RNAi were compared using two-way ANOVA with Tukey's post hoc correction.
    Could also: A generalized linear model for proportion data (e.g., logistic or quasibinomial regression) could also be applied. — Proportion/count outcomes such as hatching rate often have non-normal, heteroscedastic error structure that a GLM framework models directly, offering a complementary analysis to ANOVA on percentages.
  • Group comparisons in bar plots are summarized with mean ± SD alongside individual replicate dots.
    Could also: Reporting could also include 95% confidence intervals for the mean. — CIs directly convey the precision of the estimated mean and are often reported alongside or instead of SD, particularly useful for interpreting effect magnitude when replicate numbers are modest.
  • Agreement between the study's neuronal mRNA measurements and previously published data was summarized with a single correlation coefficient (rho) per replicate.
    Could also: A formal agreement statistic such as the intraclass correlation coefficient (ICC) or a Bland-Altman plot could also be used. — These methods are specifically designed to quantify agreement/reproducibility between two measurement sets, complementing a simple correlation coefficient.
  • Statistical significance across enrichment and differential expression analyses is largely reported as threshold categories (ns, ***, ****) rather than exact p-values.
    Could also: Exact (or exact adjusted) p-values could also be reported alongside the significance stars. — Exact values let readers judge the precise strength of evidence and support downstream re-analysis or meta-analysis of the reported comparisons.
Software: DESeq2 · GOrilla

What was reproduced

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

neuronal_STG_generation
Reported
~46 NeuroSTGs (paper's specific STG-calling pipeline, not independently reproduced)
Reproduced
mRNA Neurons WT-vs-rde4: 976 DE genes; rde4-vs-Psng1;rde-4: 1632 DE genes; sRNA Neurons WT-vs-rde4: 249 DE STGs; rde4-vs-Psng1;rde-4: 60 DE STGs (all padj<0.1)
partial
germline_STG_from_neuronal_rescue
Reported
124 germline DE genes (paper's pipeline)
Reproduced
mRNA Gonads rde4-vs-Psng1;rde-4: 205 DE genes; sRNA Gonads rde4-vs-Psng1;rde-4(SC): 1679 DE STGs
partial
sid1_independence_germline
Reported
germline STG regulation persists without sid-1
Reproduced
mRNA Gonads Psng1;rde-4 vs Psng1;rde-4;sid-1: only 63 DE genes, much smaller than other Gonads contrasts (205-3822 genes), i.e. removing sid-1 changes comparatively little on top of the rescue effect
partial
transgenerational_persistence_F3
Reported
5 genes retained differential expression at F3 (paper's specific gene-list intersection)
Reproduced
mRNA Gonads Psng1;rde-4 P0-vs-F3: 3822 DE genes; Gonads rde-4(P0)-vs-Psng1;rde-4(F3): 3800 DE genes (F3 profile still very different from rde-4 baseline); sRNA Gonads Psng1;rde-4(SC) P0-vs-F3: 192 DE STGs
partial
histone_STGs
Reported
7 histone genes among germline STG list
Reproduced
3/7 paper-listed histone genes available for spot-check via pre-curated lookup (his-46, his-61, his-63). his-61 shows large genotype-dependent normalized-count differences (rde-4 baseline ~140-307 vs rescue lines ~180-875) but padj=0.66 (not significant) in the Gonads P0-vs-F3 persistence contrast; his-46 also padj=0.66 there.
did not match
saeg2_regulation
Reported
saeg-2 downregulated by neuronal rde-4 rescue, sid-1-independent, transgenerationally relevant
Reproduced
Gonads mRNA size-factor-normalized saeg-2 counts: rde-4 baseline ~456/402/478 (high); Psng-1;rde-4 (rescue) ~42/38/35 (strongly down, ~11x); Psng-1;rde-4;sid-1 ~42/19/39 (still down, confirms sid-1-independence); Psng-1;rde-4 F3 ~166/83/310 (partial reversion, still below baseline = persists). Independently, sRNA Gonads P0-vs-F3 DESeq2 contrast shows saeg-2 antisense-siRNA signal changes significantly across generations within the rescue line (log2FC=-1.07, padj=0.079).
within tolerance

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 49/100

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: 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 🟡
Total score +9

Data is 1:1 (GSE124049, 91/91 runs, byte-exact vs ENA), and the paper's most specific claim reproduces cleanly: germline saeg-2 drops ~11x under neuronal-only RDE-4 rescue (~456/402/478 → ~42/38/35 normalized counts), stays down without sid-1 (~42/19/39), and only partially reverts at F3 (~166/83/310), corroborated independently by the small-RNA contrast (log2FC=-1.07, padj=0.079). The deviations are on our side and on the method-transparency side: the paper's bespoke STG-calling pipeline was not available, so its exact counts (46/124/40/18/7/5) were replaced by DESeq2 padj<0.1 DE-gene counts (976/205/3822), making all list-size comparisons indirect. One sub-claim genuinely did not confirm — the histone STGs, where only 3/7 genes were checkable and his-46/his-61 sit at padj=0.66 in the persistence contrast despite visible raw-count separation. A real deposition-side flaw also surfaced: 12/64 small-RNA libraries use a different 3' adapter (TruSeq Small RNA RA3), silently yielding zero reads until re-trimmed — resolvable, but undocumented in the record. Net: solid reproduction with explainable deviations, no fabrication signal, but confirmation is limited because the exact endpoints and the behavioral core claim (Fig 6) were out of reach.

🤝
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.