The Li2 mutation results in reduced subgenome expression bias in elongating fibers of allotetraploid cotton (Gossypium hirsutum L.).
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 only indirectly comparable
- 🟡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
- 🟡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 the paper's dataset (SRP026301, 4 paired-end RNA-seq libraries, byte-exact read-count match to Table 1) and its alignment pipeline (sickle trim -> GSNAP -n 1 -Q against G. raimondii D5 v2) end-to-end on «our HPC»/SLURM. Sickle trimming numbers are internally consistent and auditable. GSNAP mapping rates were obtained for all 4 libraries but only in DNA/non-spliced mode, because the -N 1 (novel splicing) flag crashes this GSNAP 2024-11-20 build with a reproducible memory-corruption bug (root-caused across 3 thread counts and 2 SIMD binary variants -- genuine upstream defect, not misconfiguration). Resulting mapping rates (74.4-82.0%) are consistently 8-10 percentage points below the paper's reported 84.4-90.2%, but preserve the EXACT SAME per-sample ranking as the paper (WT BR1 > WT BR2 > Li2 BR2 > Li2 BR1), strongly suggesting the gap is attributable to the missing splice-awareness rather than a data or reference mismatch. The paper's central biological claim (subgenome expression bias reduction under Li2) depends on a downstream homeolog-SNP-sorting + JMP Genomics statistical step that is neither in the paper's linked code repo (only 'sickle' is listed) nor reproducible with available tools -- this was NOT attempted and is explicitly out of scope for this pass, not a silent omission.
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
- 2026-07-31
- Rubric version
- v1.0
- Assessed by
-
🤖 AI curator · claude (ai-curator room) · v1.0 · run #1 2026-07-31no human curator yet
- Last updated
- 2026-07-31
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 asks how a single dominant short-fiber mutation, Ligon lintless-2 (Li2), affects homeologous (AT vs DT subgenome) gene expression in the transcriptome of elongating fibers of allotetraploid cotton (Gossypium hirsutum) compared with its near-isogenic wild type.
- ★ The Li2 mutation significantly reduces subgenome (homeolog) expression bias in the elongating fiber transcriptome. finding
- ★ The Li2 mutation has a significantly greater effect on DT-biased genes than on AT-biased genes. finding
- ★ Approximately two times more genes are differentially expressed in the AT subgenome than in the DT subgenome in both wild type and mutant fiber, i.e. the AT subgenome contributes more to the fiber transcriptome. finding
- ★ The majority of homeologous genes (83.6% of annotated G. raimondii genes at the >=10-read threshold) are expressed during fiber elongation. finding
- ★ Biological processes are unequally affected by the mutation: secondary metabolism and stress categories are significantly enriched among DT-biased genes, while photosynthesis and redox contain only AT homeologs. finding
- ★ Some homeolog pairs show reciprocal expression bias as a result of the mutation, and the direction of bias changes across developmental stages (e.g. Gorai.002G223800 switches from DT- to AT-bias in wild type but stays DT-biased in the mutant). finding
- ★ RNA-seq read mapping to the D5 (G. raimondii) reference genome combined with the PolyCat pipeline allows categorization of reads as AT-, DT-, chimeric-, or uncategorized-origin, enabling homeolog-specific expression analysis in allotetraploid cotton. method
- This is the first report exploring the effects of a single mutation on homeologous gene expression in allotetraploid cotton. finding
| Assay | System | Perturbation | Readout | Platform |
|---|---|---|---|---|
| Paired-end RNA-seq (bulk transcriptome) | Developing fibers of allotetraploid cotton Gossypium hirsutum near-isogenic lines (Li2 mutant and wild type, DP5690 background), harvested at 8 DPA (peak of elongation) | Ligon lintless-2 (Li2) dominant short-fiber mutation vs. near-isogenic wild type | Mapped read counts per gene; homeolog-specific (AT vs DT) expression levels; number of expressed and differentially expressed genes | Illumina sequencing, 101 bp paired-end reads; 4 libraries (2 biological replicates per genotype) |
| RNA-seq read mapping and homeolog categorization (bioinformatics) | RNA-seq reads mapped to the D5 reference genome of G. raimondii (37,223 genes on 13 chromosomes/pseudo-molecules) | none | Classification of reads as A-reads, D-reads, X (chimeric) reads or N (uncategorized) reads based on homeologous SNP positions; percent mapped | PolyCat pipeline |
| RT-qPCR | Developing fibers of Li2 and wild type NILs across eight developmental time points from DOA to 20 DPA (initiation, elongation, beginning of secondary cell wall biosynthesis) | Li2 mutation vs wild type | Homeolog-specific transcript levels of three genes with reciprocal expression bias (e.g. Gorai.002G223800) | — |
| Differential expression statistics (gene-by-gene ANOVA with FDR correction) | 31,114 expressed genes from Li2/WT fiber RNA-seq data | Factors: fiber type (Li2 vs wild type), AT/DT subgenome, and their combinations | Genes with FDR-corrected p<0.05 and >=2-fold difference | — |
| Functional/ontology enrichment analysis | Subgenome-biased genes (2,958 AT-biased and 1,620 DT-biased) from cotton fiber transcriptome | Li2 mutation | Relative gene frequencies per functional category; Fisher's exact test enrichment between subgenomes | MapMan ontology / BIN structure |
| Transcriptional regulator family classification | G. raimondii genome-annotated transcriptional regulators (76 families) expressed in elongating cotton fiber | Li2 mutation | Number of subgenome-biased TRs changing transcription level; fold changes per homeolog (Li2A/wtA, Li2D/wtD) | Similarity to Arabidopsis TRs (TAIR10 best hit) |
| Chi-square test of homeolog expression bias | Expressed homeologous gene pairs in wild type (29,603) and Li2 (30,842) fiber | Li2 mutation | Proportion of homeolog pairs with significant subgenome expression bias | — |
- ▼ Homeolog expression bias was reduced in Li2 fiber: 4,578/29,603 (15.5%) biased pairs in wild type vs 3,967/30,842 (12.9%) in mutant 15.5% to 12.9% (p<0.0001)
- – Li2 altered expression in both subgenomes for 35.9% (582) of DT-biased genes vs only 26.5% (784) of AT-biased genes 35.9% vs 26.5% (p<0.0001)
- ▲ About twice as many genes were differentially expressed in the AT than the DT subgenome in wild type (2,958 vs 1,620) and mutant fiber (2,574 vs 1,393) ~2-fold
- – 7,163 of 31,114 expressed genes were differentially expressed across comparisons; ~3 times fewer differentially expressed genes were found between fiber types than between subgenomes 7,163 genes; ~3-fold fewer
- – Only ~38% (583 genes of 1,536 in AT and 1,511 in DT) of genes significantly regulated between mutant and wild type overlapped between subgenomes ~38%
- – Most subgenome-biased genes decreased expression in the mutant (8.6% and 12.4%) while only a small portion increased (1.6% and 2.7%); genes down-regulated in the homeologous subgenome in wild type more often increased in the mutant (11.9% and 14% up vs 4.4% and 6.8% down) 8.6%/12.4% down vs 1.6%/2.7% up
- – Cell wall category: 12 of 40 (30%) DT-biased homeologs changed expression vs 10 of 60 (16.7%) AT-biased homeologs 30% vs 16.7%
- – Transcriptional regulators: 14 of 111 (12.6%) DT-biased vs 21 of 229 (9.2%) AT-biased TRs changed transcription level (difference not statistically significant); most decreased in the mutant 12.6% vs 9.2%
- count 639 million reads (101 bp each) from 4 libraries (Total paired-end Illumina sequencing output for Li2 and WT fiber libraries)
- other 84.4%–90.2% of reads mapped to the D5 G. raimondii reference; 29.3%–31.4% to AT and 23.4%–25.1% to DT subgenome (Read mapping and PolyCat subgenome categorization per library)
- count 34,692 of 37,223 genes (93.2%) had >=1 mapped read; 31,114 (83.6%) had >=10 reads; WT 29,603 (79.5%), Li2 30,842 (82.9%) (Expressed gene counts in developing fiber at peak elongation)
- pvalue p<0.0001 (Chi square) (Reduction of homeolog expression bias in Li2 fiber transcriptome (15.5% to 12.9%))
- pvalue p<0.0001 (Chi square) (Greater effect of Li2 on DT-biased (35.9%) than AT-biased (26.5%) genes)
- count 7,163 differentially expressed genes (FDR-corrected p<0.05, >=2-fold) (Gene-by-gene ANOVA across fiber type, subgenome and combined comparisons)
- fold_change Gorai.011G126400 (C3H(Zn)) wtA/wtD = 31.56, Li2A/wtA = 0.44, wtD/wtA = 0.03; Gorai.010G120300 wtD/wtA = 17.63, Li2D/wtD = 0.36 (Examples of subgenome-biased transcriptional regulators affected by the Li2 mutation (Table 3))
- pvalue p<0.05 (Fisher's exact test) (Enrichment of secondary metabolism and stress categories among DT-biased genes in MapMan functional analysis)
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 RNA-seq (Illumina paired-end) with two biological replicates per genotype (Li2 mutant and wild-type near-isogenic lines) to quantify homeologous gene expression in the A and D subgenomes of allotetraploid cotton fiber. Differential expression across genotype and subgenome comparisons was assessed with a gene-by-gene ANOVA using an FDR-corrected p-value threshold combined with a ≥2-fold change cutoff, while proportions of biased genes were compared with Chi-square tests and functional category enrichment was tested with Fisher's exact test. A subset of genes was validated by RT-qPCR across a developmental time course with three biological replicates, with significance indicated at p<0.05 and variability shown as standard deviation.
| Test | Applied to | n | Assumptions |
|---|---|---|---|
| Gene-by-gene ANOVA with FDR correction and ≥2-fold change cutoff | Differential expression between Li2 vs wild type, A_T/D_T subgenomes, and their combinations across 31,114 expressed genes | 2 biological replicates per genotype (4 libraries total) | not stated |
| Chi-square test | Comparison of the proportion of subgenome-biased (differentially expressed AT vs DT) genes between wild type and Li2 fiber transcriptomes | counts of expressed homeolog pairs (29,603 in WT; 30,842 in Li2) | not stated |
| Chi-square test | Comparison of Li2 mutation effect magnitude on D_T-biased vs A_T-biased genes | 2,958 A_T-biased and 1,620 D_T-biased genes in wild type | not stated |
| Fisher's exact test | Enrichment of MapMan functional categories among A_T- vs D_T-biased genes affected by the Li2 mutation | gene counts per functional category (e.g., 2,958 A_T and 1,620 D_T biased genes) | not stated |
| Significance test for RNA-seq/RT-qPCR expression differences (specific test not named in text) | A_T vs D_T subgenome expression comparison for three homeolog pairs across 8 developmental time points (Figure 3) | 2 biological replicates for RNA-seq, 3 biological replicates for RT-qPCR | not stated |
-
Differential expression across genotype/subgenome comparisons was assessed with a gene-by-gene ANOVA on RNA-seq read counts with two biological replicates per group.↳ Could also: Count-based RNA-seq models with empirical Bayes dispersion shrinkage, such as DESeq2 or edgeR — These methods are designed specifically for RNA-seq count data and can provide more stable variance estimates when biological replicate numbers are small, which may also support additional diagnostics like shrinkage-based fold-change estimates.
-
The FDR correction method applied to the gene-by-gene ANOVA p-values is not specified by name.↳ Could also: Explicitly naming the FDR procedure (e.g., Benjamini-Hochberg) or reporting q-values alongside p-values — Stating the specific multiple-testing procedure allows readers to directly reproduce the correction and compare stringency with other FDR variants (e.g., Benjamini-Yekutieli, Storey's q-value).
-
Chi-square tests were used to compare proportions of differentially expressed/biased genes between conditions.↳ Could also: Fisher's exact test or a logistic regression framework — These alternatives can also test differences in proportions and may offer more precise inference when some cell counts are small, or allow covariate adjustment within a single unified model.
-
Fisher's exact test was applied separately to multiple MapMan functional categories to test enrichment among biased genes.↳ Could also: Applying an FDR/Bonferroni correction across the set of functional categories tested — Because many categories are tested in parallel, a multiplicity correction across those tests would also help control the family-wise or false discovery rate for the enrichment analysis specifically.
-
Variability in the RT-qPCR/RNA-seq validation (Figure 3) is summarized using standard deviation from small biological replicate numbers (n=2 or n=3).↳ Could also: Reporting standard error of the mean (SEM) or a 95% confidence interval alongside or instead of SD — With very small sample sizes, SEM or CIs can communicate the precision of the estimated mean in addition to the spread of individual observations, which some readers find complementary to SD.
-
Statistical significance in Figure 3 is indicated by asterisks at p<0.05 without naming the specific test used for the subgenome comparison at each time point.↳ Could also: Reporting the specific test (e.g., paired t-test between homeolog pairs) and exact p-values — Naming the test and giving exact p-values would let readers evaluate assumptions (e.g., normality, pairing structure) and effect magnitude directly rather than relying on a threshold indicator.
What was reproduced
The exact results taken into scope, with each reported value next to the value our attempt produced.
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 a perfect match: all four SRA runs (SRR917951-954) reproduce Table 1's read counts byte-exactly (sum 638,939,992 ≈ the paper's '639 million'), so there is no cohort or sample ambiguity. The one gradeable quantitative endpoint, Table 1's mapping rate, comes out 7.9-10.0 pp low (82.0/81.6/74.4/80.6% vs 90.2/90.0/84.4/88.5%), but with the identical per-sample ranking and a uniform direction — consistent with a systematic pipeline difference on our side: the GSNAP 2024-11-20 build crashes reproducibly under -N 1 (SIGSEGV, heap corruption, confirmed across thread counts and SIMD variants), forcing non-spliced DNA-mode alignment, and our percentage uses the smaller trimmed denominator. The paper's actual headline claim — reduced AtDt subgenome expression bias under Li2 — was never tested, because the homeolog-SNP sorting pipeline is not deposited (only sickle is linked as code) and JMP Genomics is commercial; that is an authors'-side deposition gap that limits derivability, but it is a scope limitation, not evidence of a defect or fabrication. Overall: a clean, transparently reported partial reproduction with explainable deviations — yellow, not red.
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.