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

Spatial organization shapes the turnover of a bacterial transcriptome.

Elife · 2016
L1 71/100 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 🟡
Minor / cosmetic deviation
+1 pts
From: Q6 · Severity of the deviation 🟡
Concordant (toward reproduced)
Code + data deposited & functional
-2 pts
From: Data & code availability Available & functional
Total score +3
✓ What held up
  • Same input data as the authors
  • Reported values were directly comparable
  • Reported values are derivable from the shared data
  • 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
  • 🟡The deviation was non-trivial in magnitude
  • 🟡Overall, the reproduction showed a material discrepancy
How its reproducibility compares
71/100
Reproducibility score
at the mean
vs. all fields · 1173 studies
🎯 Scores higher than 38% of all assessed papers rank 694 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

PARTIAL (well-described pipeline, mostly 1:1). In-scope = the pipeline-derived mRNA-half-life computation from rifampicin tau-seq (GEO GSE75818); out-of-scope = all super-resolution STORM/smFISH imaging (Figs 1-3; raw images not deposited, matlab-storm is a generic localization toolbox). The decay model is fully specified in Methods (piecewise Eq.1 with delay/exponential/baseline; half-life=ln2/k; QC: rate error < half the fit value). I re-implemented the fit in Python/scipy on «our HPC» («job») and ran it against the paper's own deposited calibrated abundance-vs-time tables (eLife Fig-4/5/6 source data 1). RESULT: the two HEADLINE BIOLOGICAL CONCLUSIONS reproduce cleanly and independently -- (C2) inner-membrane-protein mRNAs are significantly shorter-lived than the other three localization classes which are mutually indistinguishable (reproduced both on reported half-lives, KS p=1.1e-5, and on my own refit, p=2.6e-12), and (C3) SRP-signal-peptide fusions are significantly shorter-lived than SecB/cytoplasmic controls for all 5 test genes (p same order of magnitude as reported). The underlying DECAY-RATE FIT itself (C1/C4) is well-correlated with the reported rates (Pearson 0.71-0.73, Spearman 0.79-0.81) but ~30% off in absolute value, because the exact fitting procedure (residual space, weighting, initialisation/optimiser) is not specified in Methods -- so absolute k is not byte-reproducible while ranking and all qualitative conclusions are. NO fabrication indicators: reported rates are derivable from the deposited abundances and the statistical conclusions reproduce independently. NOT ATTEMPTED (honest): (a) imaging/spatial-organization results (out of scope, raw data absent); (b) starting the pipeline from raw FASTQ -> bowtie 0.12.9 alignment -> per-CDS counts -> tmRNA(=597)/spike-in absolute calibration (the harder upstream 20%; I used the deposited calibrated abundances as fit input instead -- GSE75818 also ships per-base coverage to do this but it needs NC_000913.2 CDS annotation); (c) exact-to-the-digit KS p-values. Provisional grades; a human auditor should confirm.

💻 Code ↗ 🗄 Data: GSE75818

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 71
    assessed: 2026-06-16 ⛓ 5dd2907eb03c
✎ 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
2026-06-16
Rubric version
v1.0
Assessed by
🤖 AI curator · claude (ai-curator room) · v1.0 · run #1 2026-06-16
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: opus
Founding hypothesis

Is the bacterial transcriptome spatially organized at a genome-wide scale, and if so, what mechanism drives this organization and what consequences does it have for the post-transcriptional dynamics (degradation) of mRNAs in E. coli?

Core claims
  • The E. coli transcriptome is spatially organized genome-wide: mRNAs encoding inner-membrane proteins are enriched at the membrane, while mRNAs encoding cytoplasmic, periplasmic and outer-membrane proteins are distributed throughout the cytoplasm. finding
  • Membrane enrichment of inner-membrane-protein mRNAs is caused by co-translational insertion of signal peptides recognized by the signal-recognition particle (SRP). mechanism
  • Inner-membrane-protein mRNAs are selectively destabilized (higher degradation rates) compared with mRNAs encoding outer-membrane, cytoplasmic and periplasmic proteins. finding
  • Selective destabilization of inner-membrane-protein mRNAs is abolished when the RNA degradosome is dissociated from the membrane, implicating membrane-bound degradosomes. mechanism
  • Genomic organization (transcription site location) does not play a major role in the spatial organization of the E. coli transcriptome. finding
  • A FISH/Oligopaint method using array-derived oligo pools enables imaging of large, defined populations of mRNAs simultaneously at the transcriptome scale. method
  • mRNAs polycistronic with inner-membrane-protein messages acquire partial membrane enrichment. finding
Experimental setups
Assay System Perturbation Readout Platform
single-molecule FISH with 3D-STORM super-resolution imaging E. coli (fixed cells) none spatial distribution / density profiles of defined mRNA populations grouped by encoded-protein location and abundance 3D-STORM (stochastic optical reconstruction microscopy)
FISH with array-derived (Oligopaint) probe sets generated by enzymatic amplification E. coli antisense reverse-complement control probes vs sense probes number of single-molecule localizations per cell (labeling specificity) array-synthesized oligonucleotide pools; in vitro transcription / reverse transcription with fluorescent primer
FISH/3D-STORM of genomically-grouped transcripts E. coli none spatial distribution of mRNAs transcribed from twenty 100-kb chromosomal regions 3D-STORM
time-resolved next-generation RNA sequencing E. coli transcription inhibition (time course) to measure decay mRNA lifetimes / degradation rates grouped by encoded-protein location
time-resolved RNA-seq with degradosome-membrane-dissociation perturbation E. coli mutant dissociating RNA degradosome from membrane genetic removal of degradosome from membrane mRNA stability/degradation rates of inner-membrane vs other mRNAs
fluorescence imaging of RNA-processing enzyme localization E. coli none subcellular distribution of RNA degradosome / RNA-processing enzymes
Key results
  • Inner-membrane-protein mRNAs are strongly enriched at the cell membrane across abundance ranges
  • Cytoplasmic, periplasmic and outer-membrane-protein mRNAs are distributed throughout the cytoplasm with no strong membrane enrichment
  • mRNA-targeting probes yielded far more localizations than antisense control probes, confirming specific labeling ~10-100 fold
  • Inner-membrane-protein mRNAs have on average greater degradation rates than other mRNA groups
  • Dissociating the degradosome from the membrane preferentially stabilizes inner-membrane-protein mRNAs, abolishing selective destabilization
  • mRNAs polycistronic with inner-membrane-protein messages show partial membrane enrichment, explaining slight enrichment of periplasm/outer-membrane mRNA groups
  • Genomically grouped (100-kb region) transcripts did not show the focal organization expected if genome location dictated transcriptome organization
Key statistics
  • fold_change ~10-100 fold more localizations with sense vs antisense probes (labeling specificity control)
  • count 611 cells (average inner-membrane-protein mRNA cross-section images)
  • count 319 cells (cytoplasmic-protein mRNA distributions)
  • count 338 cells (periplasmic-protein mRNA distributions)
  • count 194 cells (outer-membrane-protein mRNA distributions)
  • count twenty 100-kb chromosomal regions (genomic-organization FISH probe sets)
  • other no single mRNA species predicted to produce more than ~10% of group signal (abundance sub-grouping design)
  • count 247 and 27 mRNA species (1/3-3 and 3-30 copies/cell) (inner-membrane-protein mRNA groups stained)

Statistical methods review

Model: opus

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 combines super-resolution imaging (3D-STORM) of FISH-labeled mRNA populations with time-resolved next-generation RNA sequencing to characterize the spatial organization and turnover of the E. coli transcriptome. Spatial distributions were reported as averages over hundreds of individual cells after computational normalization of cell dimensions, and degradation behavior was compared across mRNA groups defined by encoded-protein localization. In the portion of the text available, results are presented descriptively (e.g., average density profiles, group-level comparisons such as inner-membrane-protein mRNAs being 'on average greater' in degradation rate), with dispersion shown as SEM in at least one supplementary panel.

Replicationunclear Sample sizeImaging analyses report numbers of individual cells averaged per group (e.g., 611 inner-membrane, 319 cytoplasmic, 338 periplasmic, 194 outer-membrane cells) and numbers of mRNA species per probe group; formal sample-size or power justification is not stated in the available text. GroupsmRNA groups by encoded-protein compartment (inner-membrane, cytoplasmic, periplasmic, outer-membrane), abundance ranges, polycistronic vs not, and a degradosome-membrane-dissociation genetic perturbation Pairingna Randomization/blindingnot stated DispersionSEM
Approaches that could also have been used
  • Dispersion in at least one panel was reported as the standard error of the mean (SEM).
    Could also: The standard deviation or a 95% confidence interval could also be reported alongside or instead of SEM. — SD conveys the spread of the underlying measurements while a CI conveys precision of the estimate; reporting these (especially for smaller n) gives readers a fuller picture of variability and is often preferred for that reason.
  • Spatial organization was summarized using averaged cross-section images and density profiles built from hundreds of normalized cells per group.
    Could also: A formal statistical comparison of distributions between groups (e.g., a Kolmogorov–Smirnov test, permutation test, or mixed-effects model accounting for cell-to-cell variability) could also accompany the averaged profiles. — An explicit distributional test would add a quantitative measure of how distinguishable the group distributions are, complementing the visual/average comparison.
  • Differences in mRNA degradation between groups were described qualitatively (e.g., inner-membrane-protein mRNAs degraded 'on average greater').
    Could also: A group-level comparison of decay-rate distributions (e.g., Mann–Whitney U or a t-test on log decay rates) with reported effect sizes could also be used. — Reporting a test statistic and effect size would quantify the magnitude and statistical separation of the observed difference for readers.
  • Multiple mRNA groups (compartments, abundance ranges, polycistronic status, perturbation conditions) were compared.
    Could also: A single model framework such as ANOVA with a post-hoc correction, or a regression including these factors, could also be used. — A unified model with multiplicity control would manage family-wise or false-discovery error across the many group comparisons in one coherent analysis.
  • Time-resolved RNA-sequencing was used to estimate mRNA lifetimes/decay rates.
    Could also: Established RNA-seq differential-dynamics pipelines (e.g., DESeq2 or limma-voom with appropriate dispersion modeling and Benjamini–Hochberg FDR) could also be applied to estimate and compare decay parameters. — Such pipelines provide standardized variance modeling and built-in multiple-testing correction across the transcriptome, aiding comparability with other sequencing studies.

Result convergence & founder nodes

Findings this paper shares with others that ran a comparable experiment. A node’s strength is how many independent papers report it (replication breadth) — not how often it is cited, so a heavily-replicated but under-cited founder still stands out.

Citation network

Where this publication sits in the reproducibility-weighted citation graph — what it is built on, and what is built on it. Citation data from OpenAlex.

Citations
179
Impact: high
Foundation confidence
None of its references are in our reproducibility record yet — its foundation cannot be assessed.
Topics

No assessed neighbours yet — the network grows as more papers are assessed.

Data lineage

The datasets this paper uses (text-mined from the full text via Europe PMC), and which other assessed papers stand on the same data. A shared dataset is a factual link — not a judgement.

GSE75818 GEO in Methods (http://purl.org/orb/Methods)
no other assessed paper uses this yet
NC_000913.2 RefSeq in Methods (http://purl.org/orb/Methods)
no other assessed paper uses this yet

What was reproduced

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

Scope — pmid-27198188

Paper: Moffitt, Pandey, Boettiger, Wang, Zhuang (2016) "Spatial organization shapes the turnover of a bacterial transcriptome." eLife 5:e13065. PMCID PMC4874777. PMID 27198188.

Two experimental modalities in the paper:

  1. Super-resolution imaging (STORM / multiplexed smFISH) of the E. coli transcriptome — spatial localization of mRNAs (Figs 1–3, 7). Code link in the brief = ZhuangLab/matlab-storm (a generic STORM localization-fitting MATLAB toolbox). Raw image data are NOT in GEO and are TB-scale.
  2. τ-seq (time-resolved RNA-seq) — rifampicin-shutoff time-course RNA-seq to measure genome-wide mRNA half-lives (Figs 4–6). Data = GEO GSE75818.

In scope (pipeline-derived, attempted)

The mRNA half-life computation is the pipeline-derived result and is fully specified in Methods (eLife 13065). Pipeline: align reads (bowtie 0.12.9 → MG1655 NC_000913.2) → per-gene counts (sum over CDS) → spike-in + tmRNA(=597) calibration to copy-number-per-cell → fit decay model Eq.1 → half-life = ln2/k, with a QC filter (decay-rate error < ½ of the fit value).

Decay model (Eq.1): N(t) = N_f + N_0 · { 1 if t ≤ α exp(-k(t-α)) if t > α }

GEO ships calibrated abundance-vs-time per gene/construct in the eLife source data, and per-base coverage tables in GSE75818. Targets reproduced:

  • C1 — decay-rate fit (Fig 4 source data 1, WT). Refit Eq.1 to the paper's own calibrated abundance-vs-time and compare our decay rate to the reported decay rate, gene-by-gene. Pure reproduction of the fitting step.
  • C2 — headline biological claim (Fig 4A,B). mRNAs encoding inner-membrane proteins have significantly shorter half-lives (KS test) than cytoplasmic / periplasmic / outer-membrane mRNAs, which are mutually indistinguishable.
  • C3 — fusion-construct claim (Fig 5). SRP-signal-peptide fusions have shorter half-lives than SecB / cytoplasmic-control fusions (KS test), per test gene.
  • C4 (optional, harder 20%) — upstream counting/calibration. From GSE75818 per-base coverage (WTRep1/2) recompute per-gene counts + tmRNA calibration and check against the Fig-4 abundance columns. Needs NC_000913.2 CDS annotation.

Out of scope (not attempted; stated honestly)

  • All STORM/smFISH imaging results (Figs 1–3): localization density profiles, membrane-enrichment fractions, 611-cell averaged cross-sections. Raw images not deposited; matlab-storm is a localization toolbox, not the paper's analysis.
  • Absolute copy-number calibration from raw OD600/spike-ins starting at FASTQ (we use the deposited calibrated abundances as the fit input for C1–C3).
  • Exact reproduction of every reported KS p-value to its last digit (we reproduce the qualitative significance calls and order-of-magnitude).
Figures / tables: Fig 4Fig 4AFig 5BFig 5
C1_wt_decay_rate_fit
Reported
Per-gene WT mRNA decay rates k=ln2/half-life from rifampicin tau-seq, fit with model Eq.1 N(t)=Nf+N0*{1 if t<=a; exp(-k(t-a))}; n=10,821 gene-replicate rates reported (Fig 4 + Fig 4 source data 1)
Reproduced
Refit Eq.1 (Python/scipy) to the paper's own calibrated abundance-vs-time: Pearson r=0.725, Spearman rho=0.785 (n=10,821); median rel err 30%, 35% within 20%; QC-selection agreement 83%
partial
C2_inner_membrane_shorter_lifetimes
Reported
mRNAs encoding inner-membrane proteins have significantly shorter half-lives (KS test) than cytoplasmic/periplasmic/outer-membrane mRNAs, which are mutually statistically indistinguishable (Fig 4A,B)
Reproduced
Inner-membrane significantly shorter: KS p=1.1e-5 on reported half-lives, p=2.6e-12 on our independent refit (median 1.93 vs 2.40 min); other three groups mutually NOT significant (cyto-peri p=0.39, cyto-outer p=0.60, peri-outer p=0.78). Both significance calls + direction reproduced.
exact
C3_srp_fusions_shorter
Reported
SRP-signal-peptide fusions have significantly shorter mRNA half-lives than SecB/cytoplasmic-control fusions for all 5 test genes; SRP-vs-SecB KS p = neo 4e-20, bla 2e-22, mMaple3 2e-15, phoA 1e-21, lacZ 3e-10 (Fig 5B)
Reproduced
SRP shortest median for all 5 test genes; SRP-vs-SecB p = neo 1.1e-14, bla 4.2e-23, mMaple3 2.4e-19, phoA 5.8e-20, lacZ 1.3e-9 (all highly significant, same direction); SRP-vs-cyto also all significant. Same order of magnitude as reported.
within tolerance
C4_fusion_decay_rate_fit
Reported
Per-construct decay rates for ~24,000 signal-peptide fusion mRNAs (Eq.1; Fig 5 source data 1)
Reproduced
Refit Eq.1 to sigPep abundances: Pearson r=0.708, Spearman rho=0.813 (n=14,013); median rel err 27%
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 71/100

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 🟡
Minor / cosmetic deviation
+1 pts
From: Q6 · Severity of the deviation 🟡
Concordant (toward reproduced)
Code + data deposited & functional
-2 pts
From: Data & code availability Available & functional
Total score +3

The in-scope τ-seq half-life pipeline reproduces well: the headline conclusions (C2 inner-membrane mRNAs significantly shorter-lived with the other three classes indistinguishable; C3 SRP fusions shortest for all 5 genes) reproduce independently on a fresh refit, and reported decay rates are derivable from the deposited abundances (no fabrication). The only deviation is a ~30% absolute scatter in the decay-rate fit (C1/C4, Pearson 0.71-0.73), explained by the fit procedure being under-specified in Methods — a methodology gap on our/authors' side, not a substantive disagreement. Imaging results (Figs 1-3) are out of scope because raw images were never deposited. Overall: solid partial reproduction with explainable, non-critical deviations.

🤝
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.

154.1 k
tokens (I/O) · 10.1 M incl. cache
18 min
runtime · 0.03 CPU-h
1.9 GB
peak RAM
1
HPC jobs
hummel
machine