Targeting the epigenome and the integrated stress response to normalize colorectal cancer subclonal plasticity and progression.
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.
- ✓Reported values were directly comparable
- ✓No relevant deviation in data/preprocessing
- ✓No authors-side cause for any deviation
- ✓Reported values are derivable from the shared data
- ✓Any deviation was negligible
- ✓The central claim held under reproduction
- 🟡Could not use the authors’ exact input data
- 🟡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 the primary RNA-seq DEG claim 1:1. GEO GSE312823 ships a processed RSEM/CPM/TMM matrix (sha256 234d7c1f..., 15146 genes, 6 samples: 596-7 never-Dox x3, R1+Dox x3) that already contains the authors' own per-gene T-TEST p-value and Fold Change columns. Applying the paper's exact thresholds (P<0.01, |FC|>1.3) to those columns yields 891 DE transcripts vs the reported 925 (within-tol, 96.3%); the 3.7% gap is fully explained by 2-decimal rounding of the deposited Fold Change column right at the FC=1.3 cut (|FC|>1.27 already gives ~925). I independently recomputed the t-test from the 6 TMM-CPM replicate columns and it matches the shipped T-TEST column to ~2.5e-6 across all 15146 genes, so the statistic is genuine and the method (Student two-sample equal-variance t-test on linear TMM-CPM) is exactly identified and reproducible. NOT a different result, NOT a drop: a clean near-exact 1:1 on the checkable claim. NOT attempted (out of scope / 80-20): (a) C2's 8178-DEG contrast - the required 'Dox-induced 596-7' group was never deposited, so it is not verifiable from public data and is flagged as not-derivable-from-deposit (not asserted as fabrication); (b) IPA canonical-pathway analysis (Fig 2C, proprietary commercial tool); (c) qPCR validation (Fig 2E, wet-lab); (d) the ccbr1060 shRNA-integration WGS pipeline on SRA PRJNA1377467 (the heavy 20%). All compute on «our HPC»/«infra»; «host» holds only small results + pointers.
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 68assessed: 2026-06-14 ⛓ 849b0cabcb59
✎ 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
- 2026-06-14
- Rubric version
- v1.0
- Assessed by
-
🤖 AI curator · claude (ai-curator room) · v1.0 · run #1 2026-06-15no 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: opusThe study tests whether the integrated stress response (ISR) acts as an inducer of epigenetic and transcriptional plasticity that drives colorectal cancer subclonal diversity, treatment resistance, and progression, and whether combining epigenetic modulators with a GSPT1 degrader can normalize chromatin accessibility and reverse the resistant phenotype.
- ★ The integrated stress response induces colorectal cancer cell plasticity, subclonal diversity, and tumor progression in stress-surviving cells finding
- ★ An emergent endogenous interferon response is a key phenotypic trait associated with CRC subclonal diversity, treatment resistance, and heightened aggressiveness finding
- ★ Resistant R1 cells emerge from Dox-activated 596-7 cells by failing to express the integrated shRNA at levels sufficient to induce ISR mechanism
- ★ R1 emergence is driven by epigenetic reprogramming (altered chromatin accessibility) rather than acquisition of new oncogenic mutations mechanism
- ★ R1 displays reduced chromatin accessibility at the integrated shRNA mini-CMV promoter and at the FAS promoter, with increased repressive histone marks and CpG methylation finding
- ★ Combining epigenetic modulators with a cereblon-dependent degrader of GSPT1 normalizes chromatin accessibility and induces CRC cell death to prevent treatment-resistant progression method
- R1 is refractory to CRISPR dCas9-VP64 promoter-targeted reactivation of the integrated shRNA despite expressing comparable dCas9/sgRNA levels finding
| Assay | System | Perturbation | Readout | Platform |
|---|---|---|---|---|
| qPCR (gene/shRNA expression) | HT29-derived clonal 596-7 cells and resistant R1 subclone | Doxycycline-inducible shRNA596 (Ephrin B2 depletion) | relative shRNA/mRNA expression normalized to RPL30 | — |
| Cell death/viability assay (acridine orange/propidium iodide staining) | 596-7 cells | Doxycycline induction | percent cell death | — |
| Tumor xenograft in vivo | NOD/SCID mice bearing 596-7 (sh596) subcutaneous tumors | Dox-containing chow vs regular chow | tumor volume over time | — |
| Immunoblotting/Western blot | 596-7 and R1 cells | Dox induction / none | ISR markers (p-eIF2α, ATF3, ATF4), FAS protein | — |
| Bulk RNA-seq | 596-7 cells (±Dox) and R1+Dox cells | Dox induction | differentially expressed transcripts, pathway enrichment | — |
| FAIRE-qPCR (chromatin accessibility) | 596-7 (±Dox) and R1 (±Dox) cells | Dox / none | free DNA/total input DNA ratio at mini-CMV, FAS, IFNB1, IFNL1, ACTB, RPL30 promoters | — |
| CRISPR dCas9-VP64 transcriptional activation | HT29, 596-7, and R1 cells | mini-CMV promoter-targeting sgRNAs (#829/#830/#831) vs non-targeting (#883) | RFP fluorescence and shRNA expression | — |
| Whole-genome sequencing / bisulfite sequencing / ChIP-seq | R1 vs 596-7/HT29 cells | none | mutational burden/driver mutations, FAS CpG methylation, H3K9me3/H3K27me3/H3K4me3 marks | — |
- ▲ Dox induces ~8000-fold shRNA induction and triggers death of 596-7 cells by day 2, with viable resistant R1 colonies emerging by day 7 ~8000-fold
- ▼ Integrated shRNA expression in R1 is significantly reduced compared to Dox-induced (24h) 596-7 cells ~10-fold
- ▼ Re-introduction of Dox to Dox-deprived R1 induces shRNA re-expression at lower levels than in 596-7 ~10-fold lower
- ▼ ISR markers p-eIF2α, ATF3, ATF4 are not detected in Dox-maintained R1 na
- – RNA-seq identified 925 differentially expressed transcripts in R1+Dox vs 596-7 never-Dox and 8178 vs Dox-induced 596-7 925 and 8178 transcripts
- ▲ Interferon alpha/beta signaling predicted most significantly active pathway in R1; IFNB1/IFNL1 mRNA and protein increased in R1+Dox vs 596-7 no-Dox na
- – R1 retains all canonical HT29 driver mutations (APC, TP53, BRAF, PIK3CA, SMAD4) with similar mutational burden and no oncogenic new mutations 935 vs 942 mutations; 595 common; 59 unique
- ▼ R1 shows reduced chromatin accessibility at mini-CMV and FAS promoters with increased FAS CpG methylation and repressive histone marks 4 CpG sites within 800 bp
- fold_change ~8000-fold induction (Dox-induced shRNA activation in 596-7 cells)
- fold_change ~10-fold reduction (integrated shRNA expression in R1 vs Dox-induced 596-7)
- count 925 differentially expressed transcripts (P<0.01, FC>1.3) (R1+Dox vs 596-7 never-Dox by RNA-seq)
- count 8178 differentially expressed transcripts (P<0.01, FC>1.3) (R1+Dox vs Dox-induced 596-7 by RNA-seq)
- count 11 of 516 IPA canonical pathways met P<0.01 and Z-score ≥2 (pathways distinguishing R1+Dox from 596-7 never-Dox)
- count 17 genes (interferon/viral-genome-replication genes expressed higher in R1 vs 596-7)
- count 935 vs 942 mutations; 595 common; 59 unique to R1; 18 confirmed somatic (11 coding, 7 splice) (mutational burden R1 vs 596-7)
- count 4 CpG sites with increased methylation within 800 bp of FAS TSS (bisulfite sequencing R1+Dox vs 596-7 no Dox)
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 used a combination of qPCR, RNA-seq, chromatin accessibility assays (FAIRE-qPCR, ChIP-seq, bisulfite sequencing), and in vivo mouse xenograft experiments to characterize stress-induced epigenetic reprogramming in colorectal cancer cells. Pairwise group comparisons were primarily performed with unpaired Student's t-tests, while pathway enrichment was assessed via Fisher's exact test within Ingenuity Pathway Analysis (IPA). Results were generally presented as means ± SD from triplicate measurements, with representative experiments indicated throughout.
| Test | Applied to | n | Assumptions |
|---|---|---|---|
| unpaired Student's t-test | shRNA expression comparisons (Fig 1E); IFNB1/IFNL1 mRNA and protein comparisons (Fig 2E, F); RFP and shRNA expression after sgRNA induction (Fig 3D, E); FAIRE-qPCR chromatin accessibility (Fig 4B, E, F) | typically triplicate measurements per condition; n=5 experiments for Fig 1B cell death assay | not stated |
| Fisher's exact test | IPA canonical pathway enrichment analysis distinguishing R1+Dox from 596-7 cells (Fig 2C) | genes meeting P < 0.05, fold change > 1.3 from RNA-seq | not stated |
| RNA-seq differential expression analysis (specific test/tool not named) | Transcriptome comparisons: R1+Dox vs 596-7 never-Dox (Fig 2A); R1+Dox vs 596-7+Dox (Fig 2B) | n=3 per group (Fig 2A); n=3 R1+Dox vs n=4 596-7+Dox (Fig 2B) | not stated |
| Gene Ontology biological process enrichment (method not specified) | Pathway enrichment of DEGs in R1 vs 596-7 cells (Supplementary Fig S2B) | DEGs at P < 0.05, fold change > 1.3 | not stated |
-
Differential gene expression was reported using P < 0.01 and fold-change > 1.3 thresholds, but the specific RNA-seq statistical tool or pipeline was not named↳ Could also: Standard named tools such as DESeq2 or edgeR with Benjamini-Hochberg FDR correction (adjusted P < 0.05) are widely used alternatives — Naming the tool and reporting FDR-adjusted p-values allows readers to determine whether thresholds are applied to raw or corrected values and to reproduce the analysis
-
Multiple pairwise unpaired t-tests were used across figures with no stated multiplicity correction↳ Could also: One-way ANOVA followed by a post-hoc test (e.g., Dunnett's test against a single reference group, or Tukey HSD for all pairwise comparisons) could also be applied when three or more groups are compared — A single ANOVA with a planned post-hoc correction explicitly controls the family-wise error rate across comparisons within each experiment
-
qPCR and FAIRE-qPCR results were based on triplicate measurements from a representative experiment, with each experiment performed three times↳ Could also: Combining data across all three biological experiments and analyzing with a mixed-effects model, or aggregating experiment-level means and treating each experiment as the unit of analysis, could also be done — Modeling or pooling across biological replicates directly estimates biological variability, which is the relevant source of uncertainty for generalizing findings beyond a single run
-
Pathway enrichment was assessed using IPA's Fisher's exact test with threshold-filtered input genes (P < 0.05, fold change > 1.3)↳ Could also: Gene Set Enrichment Analysis (GSEA) on the full ranked gene list could also be applied — GSEA uses all expressed genes ranked by a continuous statistic (e.g., log2 fold change × -log10 p) rather than a binary cutoff, which can capture consistent but modest pathway-level shifts and is less sensitive to the choice of fold-change threshold
-
P-values were reported as threshold symbols (*, **, ***) rather than exact numeric values↳ Could also: Exact p-values (e.g., P = 0.004) could also be reported alongside or instead of symbols — Exact p-values enable meta-analyses, allow readers to assess the gradient of evidence, and are increasingly requested by journals and statistical reporting guidelines
-
Dispersion was reported exclusively as SD throughout↳ Could also: 95% confidence intervals could also accompany means, particularly for primary outcome measures — For small n (typically n=3 technical replicates), 95% CIs convey both spread and inferential precision simultaneously, complementing SD in communicating the uncertainty around the estimate
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.
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.
What was reproduced
The exact results taken into scope, with each reported value next to the value our attempt produced.
Scope — pmid-41963303
Paper: Targeting the epigenome and the integrated stress response to normalize colorectal cancer subclonal plasticity and progression. Cell Death Dis 2026. PMID 41963303 · PMCID PMC13181133 · DOI 10.1038/s41419-026-08720-2.
Code: https://github.com/kopardev/ccbr1060 (authors' own: "Lack J" + kopardev =
Vishal Koparde, NCI CCBR). Repo is a Snakemake shRNA-integration / chimera
detection pipeline (last push 2022-02-14). Note: the repo ccbr1060 covers the
shRNA-integration WGS analysis, NOT the bulk RNA-seq DE; the RNA-seq processed
matrix on GEO is tagged ccbr886 (a different CCBR project number).
Data: GEO GSE312823 ("Colon cancer resistance to Integrated Stress Response at a clonal level"). 6 bulk RNA-seq samples, HT-29-derived clones:
- 596-7 No Dox ×3 (GSM9354975-77)
- R1 resistant + Dox ×3 (GSM9354978-80)
Platform GPL20301 (Illumina HiSeq 4000). One supplementary processed file:
GSE312823_ccbr886_bulk_RSEM_CPM_TMM.txt.gz(RSEM quant, CPM, TMM-normalized → limma-voom-style pipeline). Also WGS fastq at SRA PRJNA1377467 (out of scope: heavy shRNA-integration re-analysis = the hard 20%).
Reported pipeline-derived results (candidate claims)
| id | reported value | paper location | in scope? |
|---|---|---|---|
| C1 | 925 differentially expressed transcripts (P<0.01, fold change >1.3) in R1+Dox vs 596-7 never-Dox | Fig 2A-B / Results | YES — both groups deposited in GEO |
| C2 | 8178 DE transcripts (P<0.01, FC>1.3) in R1+Dox vs Dox-induced 596-7 | Fig 2A-B / Results | CONDITIONAL — the "Dox-induced 596-7" group is NOT among the 6 deposited GSMs; only reproducible if the processed matrix contains extra columns for it (to be checked) |
| C3 | IPA: of 516 canonical pathways, 11 met P<0.01 & | Z | ≥2; "Interferon alpha/beta signaling" most active |
| C4 | IFNB1 / IFNL1 mRNA (qPCR validation) | Fig 2E | NO — wet-lab qPCR, not pipeline-derived |
Reproduction plan (80/20)
Primary target = C1. The shipped data is an already-normalized TMM-CPM matrix (not raw counts), so the faithful, low-cost reproduction is: run a standard limma differential-expression test on the log2(CPM) matrix for the R1+Dox vs 596-7-NoDox contrast, apply the paper's thresholds (P<0.01, fold change >1.3 → |log2FC|>0.3785), and count DE transcripts. Compare the count to 925.
Honest caveats (recorded up front):
- The authors most likely ran limma-voom on raw RSEM counts; we only have the TMM-CPM matrix → limma-trend on logCPM is an approximation of the same pipeline, not a byte-identical rerun. Exact DEG count is not expected to match to the unit.
- "P<0.01" is taken as the nominal (unadjusted) p-value (consistent with the large DEG counts reported); we also report the adjusted-p count for transparency.
- C2 depends on data that may not be deposited → flagged as possible data- incompleteness, recorded honestly rather than fabricated.
Out of scope (the hard 20%, not attempted): re-alignment of RNA-seq FASTQ with RSEM from scratch; the ccbr1060 shRNA-integration WGS pipeline (PRJNA1377467); IPA pathway analysis; wet-lab qPCR/flow/imaging.
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.
C1 (the central reproducible claim) reproduces essentially 1:1: applying the paper's stated thresholds (P<0.01, |FC|>1.3) to the authors' own deposited T-TEST and Fold Change columns gives 891 vs the reported 925, a 3.7% gap fully explained by 2-decimal FC rounding at the cutoff, with the underlying t-test independently reproduced to median |diff| 2.5e-6 across all 15,146 genes — values are genuine. The deviation is on the input/rounding side, technical and expected, not authors' fault. C2 (8178 DEGs) is unverifiable because the 'Dox-induced 596-7' comparator group was never deposited — a data-availability gap, explicitly not a fabrication signal. Overall a solid reproduction with one large but legitimately unverifiable contrast, hence yellow rather than green.
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.