Neuronal Small RNAs Control Behavior Transgenerationally.
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”.
Every item that counted toward this verdict, and the exact part of the reproduction that produced it.
- ✓Same input data as the authors
- 🟡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
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.
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 2026-07-31no 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: opusCan 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?
- ★ 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
| 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 | — |
- – 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
- 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: sonnetA 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.
| 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 |
-
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.
What was reproduced
The exact results taken into scope, with each reported value next to the value our attempt produced.
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.
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.
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.