Extensive tissue-specific expression variation and novel regulators underlying circadian behavior.
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.
- ✓Same input data as the authors
- ✓Reported values were directly comparable
- ✓The central claim held under reproduction
- 🔴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
- 🟡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 to reproduce DOWNSTREAM results faithfully. The paper's processed matrices are deposited on GEO (GSE126018), so we reproduced pipeline-derived results directly on them rather than re-running BRB-seqTools/STAR/HTSeq on 778 SRA runs. EXACT 1:1: expressed-genes-per-sample (6915/13131/9933). EXACT (qualitative): all 7 canonical clock genes cycle in all 4 tissues (JTK_CYCLE), and DGRP-796 loses ~half its rhythmic transcriptome in both brain and gut (Fig 5F). WITHIN-TOL: Fig5F gut counts (358 vs 362; 680 vs 717). PARTIAL/MISMATCH: absolute per-tissue cycling-gene tallies (135/359/479/440) come out ~1.5-2.4x higher with MetaCycle's JTK_CYCLE v3.1 at ADJ.P<0.05 single-24h period - same method, same rank-order (gut>MT>fatbody>brain), same magnitude, but exact JTK settings/extra filtering aren't pinned by the Methods, and the reported per-tissue sum (1413) contradicts the stated total (1757), so those reported anchors need human checking. NOT attempted: 45-DGRP-lines aberrant-circadian (32.1%), CircaCompare phase shift, GENIE3 network. Dataset GSE126018 profiled: open, complete, N=778 matches exactly, grade A. Overall: honest partial reproduction - headline statistic and central biology reproduce; exact cycling counts limited by under-specified parameters.
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 63assessed: 2026-06-18 ⛓ f2134371cff7
✎ 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
- 2026-06-18
- Rubric version
- v1.0
- Assessed by
-
🤖 AI curator · claude (ai-curator room) · v1.0 · run #1 2026-06-18no 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: opusHow does natural genetic variation affect circadian rhythms at the molecular level, and how does the circadian clock control tissue-specific gene expression rhythms across genotypes? The paper tests this by generating tissue-specific transcriptomes of D. melanogaster to characterize population-level molecular circadian clock variation.
- ★ More than 1700 mostly tissue-specific cycling genes are detected across four Drosophila tissues, substantially expanding the catalog of clock-controlled genes finding
- ★ Seven previously uncharacterized genes (CG2277, CG5793, CG31324, CG14688, Gclm, Amph, Usp1) cycle in all tissues and represent novel putative circadian regulators finding
- ★ Over 30% (45/141) of sampled DGRP lines exhibit aberrant, tissue-specific circadian gene expression, revealing intertissue circadian expression desynchrony driven by genetic variation finding
- ★ A novel cry mutation in DGRP-796 disrupts light-driven FAD cofactor photoreduction, providing in vivo support for the conserved photoentrainment mechanism mechanism
- ★ A regulatory TF hierarchy with tissue-specific cycling TFs acts downstream of master clock regulators to generate local gene expression rhythms mechanism
- ★ Glutathione metabolism is under tissue-specific, desynchronized circadian control, the first reported naturally occurring tissue-specific circadian molecular desynchrony of a biological process in D. melanogaster finding
- ★ A comprehensive tissue- and genotype-specific circadian gene expression atlas (>700 transcriptomes) was generated as a community resource resource
- RNAi knockdown of Amph, CG2277, CG5793, and Usp1 in clock neurons produces consistent locomotor rhythm defects, validating their role in the pacemaker finding
| Assay | System | Perturbation | Readout | Platform |
|---|---|---|---|---|
| bulk RNA-seq (BRB-seq, 3'end counting) | D. melanogaster w1118 reference strain; brain, gut, Malpighian tubules, fat body | none (12h/12h LD followed by 24h DD at 25°C) | genome-wide gene expression periodicity/rhythmicity (cycling genes) | BRB-seq |
| bulk RNA-seq (high temporal resolution RNA-seq) | 141 DGRP lines; three tissues | none (genetic variation across lines) | circadian gene expression / aberrant cycling | — |
| locomotor activity rhythm assay (behavioral perturbation) | Drosophila clock cells (tim-GAL4) and Pdf-expressing LNv central clock neurons (Pdf-GAL4) | UAS-RNAi knockdown of candidate genes | % rhythmic flies, period, rhythmicity strength/index, activity levels | UAS-RNAi constructs (Bloomington and VDRC stocks) |
| gene regulatory network / ChIP-seq integration | Drosophila cycling genes across four tissues; Clk/Cyc binding data | none | network connectivity, direct Clk/Cyc targets, TF modules | Cytoscape; publicly available Clk ChIP-seq data |
| intertissue phase shift analysis (CircaCompare) | genes cycling in two or more Drosophila tissues | none | phase shift in expression between tissue pairs | CircaCompare |
| protein structural modeling and simulation | Cry protein (DGRP-796 mutation) | novel cry mutation | effect on light-driven FAD cofactor photoreduction | — |
| brain immunohistochemistry | DGRP-796 Drosophila brain | novel cry mutation | clock function in brain | — |
| comparative ortholog circadian expression analysis | mouse (Mus musculus) and baboon (Papio anubis) tissues | none | cycling of mammalian orthologs across tissues | public databases / prior publications |
- – 1757 cycling genes detected across four tissues, most cycling tissue-specifically 1757 genes
- – Only 14 genes cycled in all four tissues, including 7 canonical clock genes and 7 uncharacterized genes 14 genes
- – 45 of 141 (>30%) DGRP lines showed aberrant circadian gene expression >30% (45/141)
- – 344 genes (19.6%) cycled in two or more tissues; 101 genes (29%) showed phase shift >2 hours between tissues 344 genes (19.6%); 101 genes (29%)
- – Glutathione-associated genes were delayed in fat body relative to gut/Malpighian tubules phase shift 4.5 to 7.8 hours
- – Knockdown of Amph, CG2277, and Usp1 increased oscillation period and decreased rhythmicity index period 23.8–25.1 hours
- – 52% of TSC genes directly connected to core regulators (Clk, cwo, vri, Pdp1); 20% directly bound by Clk/Cyc 52%; 20%
- – Most tissue-specific cycling TFs (37 of 38) showed tissue-specific cycling but not tissue-specific expression 37 of 38
- count 233 transcriptomes (w1118 reference) (four tissues profiled every 2 hours over 48 hours in triplicate)
- count 451 transcriptomes (DGRP screen) (141 DGRP lines across three tissues)
- count 1757 cycling genes (detected across four tissues by JTK_CYCLE)
- fold_change 45 (>30%) of 141 DGRP lines (lines with aberrant circadian gene expression)
- pvalue P = 5.1 × 10^−5 (GO enrichment glutathione metabolic process (Fisher's exact test))
- pvalue P = 2.2 × 10^−3 (GO enrichment glutathione transferase activity (Fisher's exact test))
- count 101 genes, 29%, FDR < 0.05 (phase shift >2 hours between at least one pair of tissues)
- other phase shift 4.5 to 7.8 hours (glutathione gene delay in fat body vs gut/Malpighian tubules)
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 generated >700 tissue-specific transcriptomes from Drosophila melanogaster (w1118 reference strain sampled every 2 hours over 48 hours in triplicate, plus 141 DGRP lines) and identified cycling genes using JTK CYCLE. Intertissue phase shifts were assessed with CircaCompare; GO enrichment was tested with Fisher's exact test; and knockdown locomotor behavior was evaluated using a test of equal proportions (% rhythmic flies) and ANOVA (period length and rhythmicity index). Multiple-testing correction was applied throughout and reported as FDR-adjusted P values with significance indicated by star notation.
| Test | Applied to | n | Assumptions |
|---|---|---|---|
| JTK CYCLE (nonparametric rhythmicity detection) | Identification of cycling genes across four tissues (brain, gut, Malpighian tubules, fat body) in w1118 reference time series | 233 transcriptomes from w1118 reference strain (4 tissues × 2-hour sampling × 48 hours × triplicates) | not stated |
| CircaCompare (phase shift comparison) | Intertissue phase shift analysis of genes cycling in two or more tissues | 344 genes cycling in ≥2 tissues | not stated |
| Fisher's exact test | GO enrichment analysis of phase-shifted genes (glutathione metabolic process P=5.1×10⁻⁵; glutathione transferase activity P=2.2×10⁻³) | — | not stated |
| Test of equal proportions | Comparison of % rhythmic flies between RNAi knockdown lines and controls in locomotor behavior assays (Fig. 2) | Per-gene N reported in Fig. 2 column | not stated |
| ANOVA | Comparison of locomotor period length and rhythmicity index across RNAi knockdown lines and controls (Fig. 2) | Per-gene N reported in Fig. 2 column | not stated |
| FDR-adjusted significance testing (specific algorithm not named) | GO enrichment of TSC genes (adjusted P values reported); intertissue phase shift analysis (FDR<0.05 threshold for 101 genes); locomotor behavior comparisons (FDR-adjusted P in Fig. 2) | — | not stated |
-
JTK CYCLE alone was used to classify genes as rhythmically expressed across tissues↳ Could also: RAIN (Rhythmicity Analysis Incorporating Non-parametric methods) or harmonic/cosinor regression (e.g., via R packages HarmonicRegress or rain) could also be applied to the same time-series RNA-seq data — Different rhythmicity algorithms differ in sensitivity and specificity depending on time-series length and sampling regularity; reporting concordant hits across two methods is an approach used in comparable circadian transcriptomics studies to characterize the robustness of the cycling gene list
-
ANOVA was used to compare period length and rhythmicity index across multiple RNAi lines and controls, with FDR adjustment applied↳ Could also: A one-way ANOVA followed by a named post-hoc procedure — Dunnett's test (each knockdown vs. a single control) or Tukey HSD (all pairwise) — could also be applied — Named post-hoc procedures yield per-comparison adjusted P values and make explicit which specific RNAi lines differ from control, which aids interpretation when several lines are tested per gene with two different drivers
-
FDR correction was applied across multiple test families, but the specific algorithm was not named↳ Could also: Explicitly naming the FDR procedure (e.g., Benjamini-Hochberg 1995 or Storey q-value) and reporting q-values alongside raw P values would also be standard practice — BH-FDR and the Storey q-value differ in conservatism and power; naming the method lets readers assess stringency and reproduce the threshold-setting step exactly
-
Fisher's exact test was used for GO enrichment of phase-shifted and TSC gene sets↳ Could also: Permutation-based enrichment testing or a topology-aware GO analysis (e.g., topGO 'weight' or 'elim' algorithms in R) could also be applied — Fisher's exact test treats each GO term independently; topology-aware methods account for the hierarchical dependency structure of the GO DAG and can reduce redundancy among enriched ancestor and descendant terms
-
The percentage of rhythmic flies was compared between knockdown and control groups using a test of equal proportions↳ Could also: A chi-square test or logistic regression with group as predictor and rhythmicity as the binary outcome could also be applied — Logistic regression additionally accommodates covariates (e.g., RNAi construct identity, driver line) and provides odds ratios as a quantified effect size alongside significance, which complements the star-notation summary shown in Fig. 2
-
Sample size for the w1118 reference time series was set to triplicates per time point without a stated power analysis↳ Could also: An a priori power calculation for time-series rhythmicity detection — for example via simulation or tools such as the R package SSPA — could also be reported alongside the design — Stating the expected effect size (e.g., amplitude of oscillation), variance, and resulting power at the chosen FDR threshold helps readers evaluate whether the design was sized to detect low-amplitude cycling genes, which are often the biologically relevant but harder-to-detect class
What was reproduced
The exact results taken into scope, with each reported value next to the value our attempt produced.
Scope — pmid-33514540 (Litovchenko et al., Sci Adv 2021; GSE126018)
In scope (pipeline-derived, attempted)
- Expressed-genes summary (count>0 per sample) from the deposited raw count matrix. Pipeline: simple matrix summary. → reproduced EXACT.
- Cycling-gene detection per tissue via JTK_CYCLE v3.1 (Methods explicitly
names it) on the deposited voom+ComBat normalized per-tissue matrices.
Tool used: MetaCycle 1.2.0
meta2d(cycMethod="JTK"), which bundles JTK_CYCLE v3.1.- w1118 reference, 4 tissues (brain, gut, fat body, Malpighian tubules).
- Fig 5F matched w1118 vs DGRP-796 (brain, gut) from the combined matrices.
- Core (intersection across 4 tissues) + main clock-gene identity.
In scope but NOT attempted this pass (harder / different data slice)
- 45 DGRP lines (32.1%) with aberrant circadian / >3.4h physiological-time shift — requires the 141-DGRP single-timepoint dataset + the authors' per-line phase estimation procedure (predict ZT from expression, compare to harvest time).
- CircaCompare v0.1.0 phase-shift (101 genes, 29%) between LD and DD — separate tool/analysis.
- Dynamic GENIE3 regulatory network / Clk-Cyc target overlap (Fig 4) — network inference, separate pipeline.
Out of scope (not a pipeline reproduction)
- Raw-read pre-processing from FASTQ (BRB-seqTools demux + STAR v2.5.0b dm3 + HTSeq v0.6.1): the authors already deposited the resulting raw count and normalized matrices, so re-running alignment on 778 SRA runs would re-derive inputs that are already provided. We reproduce DOWNSTREAM of those matrices.
- Wet-lab / behavioral assays, the cry deletion characterization (manual/molecular).
Note on code (P16)
The named repo (DeplanckeLab/BRB-seqTools) is a third-party preprocessing tool, not the downstream analysis code. Per brief P16, applying the named methods/tools (JTK_CYCLE v3.1) to the paper's own deposited data is an equally valid reproduction.
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.
Input data is open and genuine — the expressed-genes statistic reproduces exactly (6915/13131/9933.3) and Fig5F gut counts match within ~5% (358 vs 362; 680 vs 717). The core conclusions reproduce: all 7 clock genes cycle in all 4 tissues, tissue rank-order is identical, and DGRP-796 loses ~half its rhythmic transcriptome. The only substantive deviation is in absolute per-tissue cycling counts (~1.5-2.4x high), which sits on the authors' side: the JTK period window/threshold is underspecified and the reported total (1757) contradicts the per-tissue sum (1413). Not fabrication-suspect — direction and magnitude hold — so this is a solid partial reproduction with explainable, parameter-driven deviations.
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.