The oncogene AAMDC links PI3K-AKT-mTOR signaling with metabolic reprograming in estrogen receptor-positive breast cancer.
The main results reproduced: recomputed values matched the published ones within tolerance.
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
- ✓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 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 (within-tol). The paper's in-scope pipeline-derived results reproduce cleanly on a fresh «our HPC» run. The Salmon->tximport->DESeq2 differential-expression pipeline regenerates all 6 shipped DESeq2 contrasts from the shipped tximport count matrices: padj rank-correlation 1.0 and significant-gene sets (padj<0.01) IDENTICAL (Jaccard 1.0) for every contrast (5795/3477/2888/2934 drug-vs-DMSO; 888 shRNA-vs-EV; 8 EV-vs-WT); LFC Pearson 0.998-0.9997 with differences confined to low-count genes (DESeq2 1.42 vs authors' version). Fig 5b was reproduced 1:1 by re-running the authors' own correlation code on the shipped DESeq2 outputs (18025 genes), and confirms the paper's qualitative claim that dactolisib shows the strongest concordance with AAMDC knockdown. No fabrication concern: every shipped result is derivable from the shipped inputs. NOT attempted: the GEO survival analyses (no code shipped; datasets profiled instead) and the DepMap/CERES dependency figures (large figshare downloads).
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 80assessed: 2026-06-20 ⛓ e3d7201c46d6
✎ 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
- 2026-06-23
- Rubric version
- v1.0
- Assessed by
-
🤖 AI curator · claude (ai-curator room) · v1.0 · run #1 2026-06-20no 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: sonnetAAMDC, a gene centered in the 11q13.5-14.1 amplicon frequent in the poor-prognosis IntClust2 subtype of ER-positive breast cancer, acts as an oncogene that drives tumorigenesis by activating PI3K-AKT-mTOR signaling and metabolic reprogramming.
- ★ AAMDC is an amplified/overexpressed oncogene in a subgroup of ER-positive breast cancers (IntClust2) associated with poor prognosis finding
- ★ AAMDC controls PI3K-AKT-mTOR signaling, regulating translation of ATF4 and MYC and AAMDC-dependent promoter activity mechanism
- ★ AAMDC regulates expression of metabolic enzymes in one-carbon folate/methionine cycles and lipid metabolism finding
- ★ High AAMDC expression sensitizes cells to PI3K-mTOR inhibitors dactolisib and everolimus, which synergize with anti-estrogens in IntClust2 models finding
- ★ AAMDC interacts with RabGAP1L and colocalizes with RabGAP1L and Rab7a in endolysosomes finding
- ★ AAMDC knockdown inhibits BC cell proliferation, colony formation, migration, and in vivo tumor growth finding
- ★ AAMDC acts upstream of MTHFD1L, a one-carbon metabolism enzyme, in a functional pathway mechanism
- ★ Ectopic AAMDC expression activates AKT signaling and confers estrogen-independent tumor growth finding
| Assay | System | Perturbation | Readout | Platform |
|---|---|---|---|---|
| Immunohistochemistry (IHC) | breast tissue microarray, 60 high-risk BC cases | none | AAMDC H-index (intensity x % positive cells), correlation with clinical features | Biomax TMA |
| Fluorescence in situ hybridization (FISH) | 119 luminal B breast cancer specimens | none | AAMDC amplification/chr11 polysomy | AAMDC and CEN-11 probes |
| qRT-PCR / immunocytochemistry / FISH | luminal, non-luminal, and normal-like breast cell lines (e.g. MDA-MB-134, SUM52PE, SUM44PE, T-47D, MCF-12A, HuMECs) | none | AAMDC mRNA/protein expression and amplification status | — |
| shRNA knockdown with qRT-PCR and immunoblotting | SUM52PE, MDA-MB-134, T-47D breast cancer cell lines | AAMDC shRNA knockdown (sh1-3) vs empty vector | AAMDC mRNA and protein levels | — |
| Ki-67 immunostaining | SUM52PE, MDA-MB-134, T-47D cells | AAMDC knockdown | cell proliferation | — |
| Soft agar colony formation assay | SUM52PE, MDA-MB-134, T-47D cells | AAMDC knockdown | anchorage-independent growth (28 days) | — |
| Boyden migration chamber assay and phalloidin F-actin staining | SUM52PE, T-47D cells | AAMDC knockdown | cell migration and F-actin organization | — |
| Xenograft tumor model | T-47D cells in nude mice | AAMDC knockdown (sh2) vs empty vector | tumor volume at day 3 and day 10 | — |
- ▲ AAMDC copy number amplification occurs in ~10% of breast cancer cases across multiple genomic databases ~10%
- ▼ High AAMDC expression is associated with inferior overall survival in breast, ovarian, and lung cancers, and lower survival in tamoxifen-treated luminal B BC
- ▲ AAMDC overexpression by IHC is significantly more frequent in ER+ vs ER- breast tumors 47% vs 15%, p=0.005
- ▲ 25% of luminal B tumors show AAMDC amplification and a further 11% show chr11 polysomy, correlating with lymph node involvement 25% / 11%
- ▼ AAMDC knockdown inhibits cell proliferation, colony formation, and migration, and reduces xenograft tumor burden in vivo
- – RNA-seq shows AAMDC knockdown differentially down- and upregulates 1151 and 839 annotated genes respectively 1151 down, 839 up genes (q<0.05)
- ▼ AAMDC knockdown downregulates MTHFD1L, and MTHFD1L knockdown also inhibits proliferation without affecting AAMDC expression
- ▼ SUM52PE cells (AAMDC-amplified) show greater DepMap CRISPR dependence on AAMDC for survival than non-amplified luminal BC lines
- pvalue p=0.005 (AAMDC IHC overexpression ER+ (47%) vs ER- (15%) breast tumors)
- pvalue p=0.03 (nuclear AAMDC expression vs lymph node involvement (48% vs 24%))
- pvalue p=0.03 (nuclear AAMDC expression vs tumor size >5cm (36% vs 14%))
- count 25% amplification, 11% chr11 polysomy (FISH analysis of luminal B breast cancer specimens)
- count 1151 downregulated and 839 upregulated genes (q<0.05) (RNA-seq of SUM52PE cells after AAMDC knockdown)
- pvalue p=0.0217, p=0.0018, p<0.0001 (qRT-PCR AAMDC expression across breast cell lines vs MCF-12A)
- pvalue p=0.0003 to p<0.0001 (xenograft tumor volume, AAMDC KD vs EV, T-47D model in nude mice)
- count n=3 biologically independent experiments (in vitro proliferation, colony formation, and migration assays)
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 paper uses a mix of clinical/genomic association analyses (chi-square tests, Kaplan-Meier survival curves with log-rank tests) and in vitro/in vivo functional experiments (qRT-PCR, proliferation, colony formation, migration, xenograft tumor volume) analyzed predominantly with two-tailed unpaired Student's t-tests or one-way ANOVA with Dunnett's multiple comparison test. RNA-seq differential expression was reported using log2 fold-change with a q-value (FDR-adjusted) threshold of <0.05. Results are reported as mean ± SD or mean ± SEM depending on the figure, with asterisks denoting significance thresholds and stated biological replicate numbers (typically n=3 for cell culture experiments).
| Test | Applied to | n | Assumptions |
|---|---|---|---|
| log-rank test | Kaplan-Meier survival plots (Fig. 1b) | — | not stated |
| Chi-square test (one-tailed) | AAMDC IHC expression vs. lymph node metastasis, tumor grade, tumor size, ER status (Fig. 1d) | n = 60 biologically independent samples | not stated |
| Ordinary one-way ANOVA with Dunnett multiple comparison test | qRT-PCR AAMDC expression across luminal, non-luminal, and normal-like breast cell lines relative to MCF-12A (Fig. 1f) | n = 3 biologically independent RNA extractions | not stated |
| two-tailed unpaired Student's t-test | AAMDC knockdown validation (qRT-PCR/immunoblot), Ki-67 proliferation, soft agar colony formation, Boyden migration assays (Fig. 2a-d) | n = 3 biologically independent experiments | not stated |
| two-tailed unpaired Student's t-test | xenograft tumor volume, T-47D model (Fig. 2f) | n = 8 mice per group | not stated |
| differential expression testing with log2 fold-change and q-value thresholding | RNA-seq comparing AAMDC knockdown (sh2) vs. EV/untransduced SUM52PE cells (Fig. 3a-d, Supplementary Data 2) | — | not stated |
-
Multiple pairwise two-tailed unpaired t-tests were used across several shRNA conditions (EV, sh1, sh2, sh3) within the same experiment (Fig. 2a-d) without a stated correction for the resulting multiple comparisons.↳ Could also: A one-way ANOVA with a post-hoc multiple comparisons correction (e.g., Tukey's HSD or Dunnett's test, as was already used in Fig. 1f) — This approach would jointly control the family-wise error rate across all pairwise comparisons within a single experiment, which can be a useful complement to per-comparison t-tests when several groups are compared to each other or to a common control.
-
Dispersion is reported as SD in some figures (e.g., Fig. 2a-e) and as SEM in others (e.g., Fig. 1f, Fig. 2f), with small sample sizes (often n=3).↳ Could also: Reporting a 95% confidence interval alongside or instead of SD/SEM — A CI directly conveys the precision of the estimated effect and can be more intuitive for readers than SEM, which shrinks with sample size and can visually understate variability at small n.
-
The chi-square analysis in Fig. 1d used a one-tailed p-value to relate AAMDC localization to clinical features in a sample of n=60.↳ Could also: A two-tailed chi-square test, or Fisher's exact test for smaller subgroup cell counts — A two-tailed test does not presuppose the direction of association in advance, and Fisher's exact test is often preferred over chi-square when contingency table cell counts are small, providing exact rather than asymptotic p-values.
-
Functional assays (proliferation, colony formation, migration) relied on n=3 biological replicates analyzed with a parametric t-test.↳ Could also: A non-parametric test such as the Mann-Whitney U test — With very small sample sizes, normality is difficult to assess or verify; a non-parametric alternative does not rely on a normality assumption and can be a useful complementary or alternative approach.
-
Survival differences were assessed using the log-rank test on Kaplan-Meier curves stratified by AAMDC expression level.↳ Could also: A Cox proportional hazards regression model — A Cox model can incorporate additional clinical covariates (e.g., tumor stage, treatment) alongside AAMDC expression, yielding an adjusted hazard ratio and confidence interval rather than only a p-value for the unadjusted comparison.
-
RNA-seq differential expression was summarized using log2 fold-change and a q-value cutoff, without the specific software/package or normalization method stated in the excerpted text.↳ Could also: Standard RNA-seq pipelines such as DESeq2 or edgeR with explicit reporting of normalization method and multiple-testing procedure (e.g., Benjamini-Hochberg FDR) — Explicitly naming the differential expression package, version, and FDR method supports reproducibility and allows readers to evaluate the specific statistical model (e.g., negative binomial dispersion estimation) underlying the reported q-values.
What was reproduced
The exact results taken into scope, with each reported value next to the value our attempt produced.
Scope — PMID 33772001 (Golden et al. 2021, Nat Commun)
Paper: "The oncogene AAMDC links PI3K-AKT-mTOR signaling with metabolic reprograming in estrogen receptor-positive breast cancer." Nat Commun 12:1920. DOI 10.1038/s41467-021-22101-7. PMCID PMC7998036.
Repo: https://github.com/jcursons/Golden_2021_NatComm (author code; J. Cursons). Cloned to «infra» reproductions/pmid-33772001/repo/.
What the repo actually ships (self-contained)
data/Salmon_tximport_geneLevel-DrugInhib.csv— tximport gene-level output (abundance+counts+length) for 15 drug-treated SUM52PE RNA-seq samples (DMSO/Dactolisib/Everolimus/AZD8055/Buparlisib ×3).data/Salmon_tximport_geneLevel-shRNA.csv— same, 9 samples (AAMDC-shRNA / EV / WT ×3).- 6 shipped DESeq2 result CSVs (
SUM52-AAMDC_*_-_DESeq2_geneLevel.csv): Dacto/Evero/AZD/Bupar vs DMSO, shRNA vs EV, EV vs WT. data/RScript/*.RScript— the exact tximport+DESeq2 code that produced those CSVs.Golden_2021_NatComm.py— figure code:drug_effects_vs_shrna_effects()→ Fig 5b (drug-vs-shRNA log2FC correlation panels), andbrca_ceres_assoc()→ DepMap dependency figures (auto-downloads ~700 MB–1 GB from figshare).data/GRCh38_98_ENSGToHGNC.pickle— ENSG→HGNC map (no GTF download needed for Fig 5b).
IN SCOPE (pipeline-derived, attempted)
- DESeq2 differential expression (6 contrasts). Re-run DESeq2 from the shipped tximport count matrices (reconstruct the
txiobject) and compare to the shipped DESeq2 result CSVs. Pipeline: Salmon→tximport→DESeq2. Deterministic. Primary claim. - Fig 5b — drug-vs-shRNA log2FC correlation. Run the authors'
drug_effects_vs_shrna_effects()on the shipped DESeq2 outputs; regenerate the 5 comparison panels and the on-panel R²/Spearman statistics. Self-contained. - (stretch) DepMap dependency figures (
brca_ceres_assoc(), Fig 2-related). Requires large figshare downloads (CERES/DEMETER2/CCLE RNA/CNV). Attempt if feasible.
OUT OF SCOPE / NOT ATTEMPTED
- GSE11121 patient survival analysis. Methods: "survival analyses of breast, ovarian, and lung cancer patients were performed by investigating the GSE11121, GSE13876, and GSE19188 GEO datasets." This is a Kaplan–Meier/survival analysis whose code is NOT in the repo — only the SUM52 RNA-seq DE pipeline + figure code are shipped. Dataset is profiled (see
data/dataset_profile.json) but the survival result cannot be reproduced from shipped code. GSE13876/GSE19188 likewise have no shipped code. - All wet-lab results (cell viability, westerns, metabolomics, xenografts, IHC, ChIP) — not pipeline-derived.
- Manual/external-tool analyses without shipped code.
Notes
- Env («our HPC»): DESeq2 1.50.2 / tximport 1.38.2 / R 4.5.3 (shipped CSVs predate this; expect tiny version-driven LFC differences for low-count genes). Fig 5b: python 3.10, numpy 1.23.5, scipy 1.15, pandas 2.3, matplotlib 3.10, adjustText.
- The accession named in the BRIEF (GSE11121) is not the repo's primary data; the repo's data is the SUM52PE RNA-seq (no GEO accession shipped in repo; deposited array data per paper). Both are profiled/assessed.
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.
The DESeq2 differential-expression core reproduces 1:1 from the authors' shipped tximport counts: 6/6 contrasts, padj Pearson r=1.0, identical significant-gene sets (Jaccard=1.0), LFC Pearson 0.998-0.9997. The near-perfect concordance is expected determinism (same input + same tool), not a too-perfect/fabrication red flag. The reproduction is preliminary and partial — Fig 5b still running, GSE11121 survival and DepMap figures out of scope — so the full central AAMDC/PI3K-AKT-mTOR conclusion is only partly testable, yielding a solid yellow rather than a clean 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.