Identification of new ETV6 modulators through a high-throughput functional screening.
The main results reproduced, with only marginal, non-material deviations.
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 reproduction (both in-scope pipeline results reproduced; marquee results out-of-scope by data availability). Reproduction unit = GEO GSE79373 + Cufflinks (third-party tool; paper reports no original code). C1 (deposit integrity): independently re-aligned both deposited Reh RNA-seq runs (REHWT SRR3236162, REH2-11Plenti SRR3236158) with STAR 2.7.10b -> HTSeq 2.0.5 (reverse-stranded, Ensembl GRCh37.75) and compared per-gene to the deposited GSE79373 HTSeq counts: Pearson(log1p) 0.9886/0.9885, ~84% exact per-gene counts, total counts within 0.7% -> deposit integrity CONFIRMED. C2 (the assigned Cufflinks step): Cufflinks v2.2.1 on the REHWT BAM yields 63653 gene FPKMs; the paper's FPKM<=0.21 'non-expressed' cutoff splits them into 46448 non-expressed (median deposited count 0, 76% zero) vs 17205 expressed (median deposited count 813, 5% zero) -> the cutoff is a valid, reproduced non-expressed threshold. OUT OF SCOPE (not deposited under any accession, recorded not graded): the headline shRNA functional screen (~140k shRNAs; 2858 positive -> 1241 retained -> 90 candidate genes -> 13 validated; top5 AKIRIN1/COMMD9/DYRK4/JUNB/SRP72) and the targeted-RNA-seq Table 1 ratios. The exact 2858->1241 shRNA reduction is not reproducible because it needs the undeposited shRNA->target mapping, but the Cufflinks/FPKM<=0.21 step it depends on does reproduce. Honest verdict: described well enough for the deposited Reh expression (1:1 on deposit integrity, valid on the FPKM cutoff); the paper's central screen rests on undeposited data and cannot be independently checked.
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
- 2026-06-25
- Rubric version
- not recorded
- Assessed by
- —
- 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: sonnetETV6's regulatory network remains unclear despite its role in hematological malignancies; the paper tests whether an unbiased genome-wide shRNA screen can identify novel genes that modulate ETV6's transcriptional repressive activity.
- ★ A genome-wide shRNA screen in an engineered ETV6-dependent Blasticidin-sensitive pre-B ALL cell line can identify modulators of ETV6 repressive transcriptional activity method
- ★ 13 shRNAs were identified that knock down their target genes and induce overexpression of ETV6 transcriptional target genes finding
- ★ Silencing of AKIRIN1, COMMD9, DYRK4, JUNB, and SRP72 leads to abrogation of ETV6 repressive activity finding
- ★ The Reh EBS3tk BlastR clone 2-1 ETV6-5 cell system is a suitable, reversible model to screen for ETV6 modulators resource
- ★ The identified modulators have a broad, gene-specific impact on the ETV6 transcriptional network finding
- ★ AKIRIN1, DYRK4, SRP72, and JUNB modulate t(12;21)-associated ETV6 target genes, suggesting conservation of modulator activity across leukemic contexts finding
- Structure-prediction-based interaction analysis can nominate additional ETV6 modulator candidates missed by single-shRNA filtering method
| Assay | System | Perturbation | Readout | Platform |
|---|---|---|---|---|
| genome-wide shRNA loss-of-function screen with deep sequencing | Reh pre-B ALL cell line (EBS3tk-BlastR, ETV6-expressing, clone 2-1 ETV6-5) | shRNA library (~140,000 unique shRNAs, 15 pools) | shRNA enrichment/normalized counts under Blasticidin selection vs control | — |
| qRT-PCR | Reh EBS3tk BlastR clones (2-1 to 2-22) | lentiviral ETV6 overexpression | BlastR mRNA expression | — |
| ChIP-seq | Reh clone 2-1 | none (endogenous ETV6 ChIP) | ETV6 binding/integration site of EBS3tk BlastR | — |
| PCR (genomic) | Reh clone 2-1 | none | confirmation of EBS3tk BlastR insertion site in NEGR1 intron 1 | — |
| qRT-PCR / Western blot | Reh 2-1 ETV6-5 clone | shRNA knockdown of ETV6 or NAM (nicotinamide) treatment | BlastR expression and ETV6 protein level | — |
| proliferation assay (cell counting) | Reh 2-1 ETV6-5, shCtl-GFP, shETV6-GFP sorted cells | Blasticidin dose range, ETV6 knockdown | cell proliferation over 12 days | — |
| targeted RNA sequencing (custom gene panel) | Reh ETV6-expressing cell lines individually transduced with candidate shRNAs (90 candidates) | shRNA knockdown of individual candidate genes | expression of 84-gene ETV6 target/control panel | — |
| qPCR validation | Reh shDYRK4 cells | shRNA knockdown of DYRK4 | DYRK4 expression (knockdown efficiency) | — |
- ▼ ETV6 re-expression reduces BlastR expression in Reh clone 2-1 compared to pLENTI control 2.7-fold, p ≤ 0.001
- – 126,042 unique shRNAs identified by deep sequencing; 1,241 retained after threshold filtering, yielding 90 candidate ETV6 modulator genes
- ▲ 13 shRNAs efficiently knocked down target genes and impaired ETV6 transcriptional activity
- ▲ 5 top modulators (AKIRIN1, COMMD9, DYRK4, JUNB, SRP72) significantly impact expression of ETV6 target genes
- ▼ 40 of 84 candidate genes confirmed as ETV6-dependent targets; 22 of these are direct ETV6 ChIP-seq targets
- ▼ shDYRK4 knockdown efficiency validated by qPCR 73% knockdown
- – AKIRIN1, DYRK4, SRP72, and JUNB knockdown samples cluster with ETV6-null pLENTI sample based on t(12;21)-associated target gene expression; COMMD9 shows only moderate impact
- – EBS3tk BlastR insertion into NEGR1 intron 1 does not alter NEGR1 expression and NEGR1 is not significantly ETV6-regulated 53.1 vs 51.3 FPKM; logFC=-0.46, FDR=0.50
- fold_change 2.7-fold reduction in BlastR expression with ETV6 vs pLENTI (qRT-PCR validation of ETV6-dependent BlastR repression in clone 2-1)
- pvalue p ≤ 0.001 (BlastR repression by ETV6 vs pLENTI control)
- count 126,042 unique integrated shRNAs out of ~140,000 library (genome-wide shRNA screen sequencing depth)
- count 2,858 positive shRNAs; 1,241 retained after filtering (shRNA screen threshold selection)
- count 81 genes targeted by ≥2 distinct shRNAs (77 by two, 4 by three) (prioritized candidate genes from screen)
- count 90 total candidate modulator genes (81 + 9 from structure prediction) (final candidate gene list for validation)
- count 40 of 84 candidate target genes confirmed ETV6-dependent; 22 direct ChIP-seq targets (targeted RNA-seq validation of ETV6 transcriptional network)
- fold_change logFoldChange = -0.46, FDR = 0.50 (NEGR1 expression change upon ETV6 re-expression)
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 genome-wide shRNA screen with sequential read-count thresholds (including a 2 SD cutoff relative to a control shRNA) to nominate ETV6-modulator candidates, then validated candidates with targeted RNA sequencing and qRT-PCR. Comparisons between conditions (e.g., BlastR expression across clones/treatments, shRNA pairs/trios, knockdown categories, and ETV6-target versus control-gene expression) were assessed with paired or unpaired Student's t-tests, and grouping patterns were additionally visualized with unsupervised hierarchical clustering/heatmaps. Results were reported mainly as mean ± SD with exact p-values (Table 1) or threshold-based significance markers (*, **, ***) in figures.
| Test | Applied to | n | Assumptions |
|---|---|---|---|
| Student's t-test (paired) | Figure 3E: comparison of normalized counts between best and second-best shRNA of a pair/trio | n = 81 gene pairs/trios | not stated |
| Student's t-test (unpaired) | Figure 4D: relative expression compared across knockdown categories (Best KD, KD, No KD) | Best KD n=31, KD n=32, No KD n=22 | not stated |
| Student's t-test | Figure 4C: p values for ETV6-target vs non-target gene expression per shRNA sample | 22 target genes vs 62 control genes per shRNA | not stated |
| Student's t-test (paired) | Table 1: 'Targets SHC vs shRNA (paired)' p-values for the five top modulators | 22 ETV6 target genes per comparison | not stated |
| Student's t-test | Figure 2A: BlastR expression in ETV6-expressing clone vs mock/pLENTI controls | n = 4 technical replicates | not stated |
| SD-based threshold for hit calling (2 SD above SHC202 control counts) | Figure 3A: selection of over-represented shRNAs in the genome-wide screen | 126,042 unique integrated shRNAs screened | not stated |
-
Technical replicate variability in qRT-PCR assays (n=4) is summarized as mean ± SD.↳ Could also: Reporting the mean with a 95% confidence interval or SEM alongside SD — A CI directly conveys the precision of the estimated mean, which can be a useful complement to SD (which describes spread of the replicate measurements themselves) when communicating how confidently a difference between conditions is estimated.
-
Multiple independent Student's t-tests were performed across many shRNA and gene-target comparisons (e.g., Figure 4C, Table 1) without a stated multiple-testing correction.↳ Could also: Applying a family-wise error control (Bonferroni, Holm) or a false-discovery-rate procedure (Benjamini-Hochberg) across the full set of comparisons — When many tests are run on the same dataset, an FDR or family-wise correction is a standard way to keep the expected proportion of false positives controlled across the whole comparison set, complementing the per-comparison p-values already reported.
-
Hit selection in the genome-wide shRNA screen used a fixed 2 SD threshold relative to a control shRNA to call over-represented shRNAs.↳ Could also: A formal screen-analysis pipeline such as MAGeCK, RSA (redundant siRNA activity), or a permutation/empirical null-based FDR approach — These established screen-analysis frameworks incorporate replicate variability and multiple-hairpin redundancy directly into a statistical hit-calling score, which can complement a simple SD-based cutoff, particularly for genome-wide libraries with many hairpins per gene.
-
Grouping of expression profiles across shRNA samples was visualized with unsupervised hierarchical clustering/heatmaps (Figures 4H, 5A).↳ Could also: Pairing the clustering with a formal group-separation statistic, such as PERMANOVA or a silhouette/bootstrap-support metric — A quantitative test of cluster separation can complement the visual heatmap pattern with a numerical measure of how distinct the identified groups are.
-
Comparisons between paired best/second-best shRNAs (Figure 3E) and paired SHC-vs-shRNA target expression (Table 1) used paired Student's t-tests.↳ Could also: A non-parametric paired alternative such as the Wilcoxon signed-rank test — A non-parametric paired test can also be used when normality of the paired differences is uncertain, and would complement the parametric t-test results, especially for smaller sample sizes.
-
Differential expression for a single gene (NEGR1) was reported with a logFoldChange and an FDR value from RNA-seq data.↳ Could also: Extending the same FDR-based differential expression framework (e.g., DESeq2, edgeR, or limma-voom) consistently to the targeted RNA-seq validation comparisons — Using the same model-based differential-expression approach across all RNA-seq-derived comparisons, rather than combining it with t-tests for some analyses, can offer a unified variance-modeling framework across the whole dataset.
What was reproduced
The exact results taken into scope, with each reported value next to the value our attempt produced.
Scope — pmid-35198911
Paper: Neveu et al. 2022, iScience — "Identification of new ETV6 modulators through a high-throughput functional screening." PMID 35198911 / PMC8851229 / DOI 10.1016/j.isci.2022.103858.
Assigned reproduction unit (from BRIEF.md):
- Code: https://github.com/cole-trapnell-lab/cufflinks (Cufflinks — third-party tool)
- Data: GEO GSE79373
Important up-front facts (control-plane verified)
- GSE79373 is REUSED data, generated for the lab's earlier paper Neveu et al. 2016, Blood "CLIC5: a novel ETV6 target gene..." (PMID 27540136). The 2022 paper's Data Availability cites GSE79373 (RNA-seq), GSE102785 (ChIP-seq), PXD031001 (proteomics).
- The 2022 paper's code-availability statement reads verbatim: "This paper
does not report original code." The assigned GitHub link (
cole-trapnell-lab/ cufflinks) is the generic third-party Cufflinks tool, NOT authors' code. Per BRIEF rule P16, applying this third-party tool to the paper's own data is a valid reproduction.
GSE79373 contents (from GEO sample table)
- 25 RNA-seq samples, Homo sapiens, genome hg19 / Ensembl GRCh37.75.
- 17 patient ALL samples (
315_T…856_T) on AB 5500 SOLiD (Lifescope align). - 7 Reh cell-line samples on Illumina HiSeq 2500 (STAR align): GSM2093552 REH2-11ETV6-ETS-NLS-HIS, GSM2093553 REH2-11ETV6-HIS, GSM2093554 REH2-11Plenti, GSM2093555 REH2-1ETV6-ETS-NLS-HIS, GSM2093556 REH2-1ETV6-HIS, GSM2093557 REH2-1Plenti, GSM2093558 REHWT.
- Deposited processed files = HTSeq counts (
*.counts.txt.gz), GSE79373_RAW.tar (~5.2 MB). - GEO data_processing: Lifescope 2.1 (SOLiD) / STAR (Illumina) → PICARD 1.107 mark-dup → GATK 3.2-2 SplitNCigar → HTSeq 0.6.1p1 counts vs Ensembl GRCh37.75.
- Raw reads: SRA SRP071966.
2022 paper bioinformatic pipeline (verbatim from Methods)
- shRNA screen: "aligned on a custom reference containing all the target sequences of the MISSION Human shRNA library using the Bowtie2 ... version 2.2.3"; "read counts per shRNA were obtained using BEDTools version 2.22.1"; normalized per total reads.
- Targeted RNA-seq: raw counts normalized per total reads + 12 housekeeping genes; relative expression vs SHC002/SHC202 controls.
- Expression filter: "shRNAs which target non-expressed genes (FPKM ≤0.21) in Reh cells."
- "The gene expression values (FPKM) were calculated using the Cufflinks tool version 2.2.1 on hard-clipped BAM files."
In scope (pipeline-derived AND backed by deposited data)
- R1 — Reproduce GSE79373 deposited Reh counts. Re-align a Reh Illumina
sample (REHWT = GSM2093558; control REH2-11Plenti = GSM2093554) from SRA reads
with STAR → HTSeq (Ensembl GRCh37.75), compare per-gene counts to the deposited
*.counts.txt.gz. Verifies the deposit reproduces. (Pipeline: STAR+HTSeq.) - R2 — Cufflinks FPKM in Reh cells + the FPKM ≤0.21 filter. Run Cufflinks v2.2.1 on the Reh BAM (the assigned tool) to compute FPKM; characterize the expressed/non-expressed split and verify that FPKM ≤0.21 is a sensible non-expressed cutoff (this is the exact 2022 step that uses GSE79373 + Cufflinks).
Out of scope (no deposited data and/or wet-lab/manual)
- Headline screen counts (1,241 shRNAs → 81 genes → 90 candidates → 13 validated; Fig 3–4, Table 1): the shRNA-barcode deep-sequencing data is not deposited under any accession → cannot reproduce (blocker: no_data_accession for that assay).
- Targeted RNA-seq relative-expression (Table 1 ratios, e.g. DYRK4 3.21, SRP72 2.55): targeted-RNA-seq reads not deposited → cannot reproduce.
- Proteomics (PXD031001), ChIP-seq (GSE102785), reporter/wet-lab assays: out of scope for this RU (different accessions / wet-lab).
Honest expected outcome
A partial reproduction: the only deposited+pipeline+named-tool intersection is the Reh-cell expression (R1/R2). The paper's marquee results are not backed by deposited data,
Assessments & scoring basis
Each contributor’s verdict, the per-question basis, and the auditable, itemised worksheet behind it.
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.