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

A 14-bp motif in the KIT active promoter region is critical for melanin accumulation in yaks, mice, and humans.

BMC Biol · 2025
L1 54/100 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: Q5 · Derivability / plausibility 🟡
Content-critical question only partially held
+2 pts
From: Q7 · Core claim 🟡
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: Q2 · Endpoint comparability 🟡
Input / endpoint not comparable 1:1
+1 pts
From: Q1 · Data identity 🔴
Concordant (toward reproduced)
Code + data deposited & functional
-2 pts
From: Data & code availability Available & functional
Total score +7
✓ What held up
  • Any deviation was negligible
What did not (or only partly)
  • 🔴Could not use the authors’ exact input data
  • 🟡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 central claim did not (fully) hold under reproduction
  • 🟡Overall, the reproduction showed a material discrepancy
How its reproducibility compares
54/100
Reproducibility score
1.1 SD below mean
vs. all fields · 1173 studies
🎯 Scores higher than 14% of all assessed papers rank 997 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 reproduction. The paper is described well enough to map its pipeline, but its key processed artifacts were NOT deposited: there is no 387-yak WY18 joint-call VCF, and 247 of 387 sample accessions are unstated (only ~140 new yaks / 149 runs are public in PRJNA1069901). So the cohort-wide numbers C1-C3 and the Fst/XP-CLR scan C6 are not reproducible from public data (not attempted, documented). The named public code is PopLDdecay, a generic third-party LD tool; the only public processed genotypes are a DIFFERENT study's bovine-pangenome SNP callset. On «our HPC» («job») we: (R2/C4) CONFIRMED the headline KIT locus on the public WY18 reference — the E2F core motif CCGCGGGAAG is present (minus strand) at exactly Chr6:72,971,725 in a GC/CpG-rich active promoter [within-tol]; (R1) built PopLDdecay at the pinned commit 425f978 and ran it on 326 yaks, but the LD-decay curve is uninterpretable because the public pangenome callset is too sparse (1684 Chr6 SNPs) [partial/methodological]; (R3/C5) as a proxy for the indel's weak LD, found 81.5% of flanking yak SNP pairs have r2<0.8 (mean 0.45) — supportive but indirect since the indel is absent from the SNP-only public set [partial]. Did NOT attempt: cohort variant counts, selection scan, GWAS/PCA/ADMIXTURE, RNA-seq DEGs, or any wet-lab claim. This is a feasibility/honesty result, not a refutation: the biological locus checks out on public data, while the cohort statistics depend on undeposited artifacts.

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.

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
2026-06-23
Rubric version
v1.0
Assessed by
🤖 AI curator · claude (ai-curator room) · v1.0 · run #1 2026-06-24
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: sonnet
Founding hypothesis

The study tests whether a specific genetic variant in the KIT gene promoter is the causal regulatory mutation responsible for reduced KIT expression and loss of melanin accumulation (all-white coat phenotype) in yaks, and whether this variant's effect is conserved in mammals including mice and humans.

Core claims
  • A 14-bp deletion in the KIT gene promoter region is associated with the all-white phenotype in yaks, identified via GWAS finding
  • The 14-bp deletion lies within an active KIT promoter region and disrupts binding of transcription factors E2F3 and E2F6, reducing KIT expression mechanism
  • CRISPR/Cas9 knock-in mice lacking the yak 14-bp promoter motif show reduced KIT expression and white coat spots, confirming functional causality finding
  • CRISPR/Cas9 deletion of the human orthologous 14-bp motif in melanoma cells (A375, A875) reduces melanin accumulation finding
  • All-white yaks cluster into two distinct genetic groups (W1 and W2) based on NJ tree, PCA, and admixture analysis finding
  • The 14-bp promoter sequence is conserved across 12 mammal species and overlaps histone modification marks (H3K4me1, H3K4me3, H3K27ac) in human melanocytes finding
  • SCF-cKIT signaling pathway genes show decreased expression in white-coat yak ear tissue compared to black-coat yaks finding
  • The 14-bp deletion evolved together with the SCs6 structural variant, jointly contributing to the all-white phenotype in yaks mechanism
Experimental setups
Assay System Perturbation Readout Platform
Whole-genome sequencing / GWAS, FST, XP-CLR 387 domestic and wild yaks (Bos grunniens, B. mutus) none (natural coat color variants compared) genetic variants (SNPs/indels), association signals, selection signatures
PCR genotyping 40 re-sampled all-white yaks none confirmation of 14-bp deletion presence PCR
Dual-luciferase reporter assay HEK293T cells promoter fragment truncation/deletion (WT vs Mut lacking 14-bp) luciferase activity as measure of promoter activity pGL3-Basic vector with pRL-TK internal reference
Dual-luciferase reporter assay with transcription factor co-transfection HEK293T cells co-transfection with E2F3 or E2F6 expression plasmids plus WT or Mut promoter luciferase activity indicating transcription factor binding pcDNA3.1 vector
RNA-seq transcriptome analysis and Western blot yak ear tissue (6 black-coat, 6 white-coat) none (natural coat color comparison) KIT and SCF-cKIT pathway gene expression (PFKM), protein levels
CRISPR/Cas9 gene knock-in C57BL/6J transgenic mice replacement of mouse KIT promoter with yak WT or 14-bp-deleted promoter coat color phenotype, KIT mRNA expression CRISPR/Cas9
CRISPR/Cas9 knockout, melanin assay, immunofluorescence, transcriptome (RNA-seq) human melanoma cell lines A375 and A875 24-bp deletion spanning human 14-bp orthologue (hKIT Pro1) vs controls (Pro2, mock) melanin content, pigment granule localization, differentially expressed genes CRISPR/Cas9 lentivirus vector
Key results
  • 14-bp deletion at Chr6:72,971,725 is homozygous absent in 20.0% and heterozygous absent in 80.0% of all-white yaks, present in other yaks 20.0%/80.0%
  • WT promoter construct induced significantly higher luciferase activity than Mut (14-bp deleted) construct p<0.01
  • E2F3 and E2F6 co-transfection stimulated expression more with WT plasmid than with Mut plasmid
  • Homozygous ymKIT Pro14Del/Pro14Del mice showed white spots on abdomen and significantly decreased KIT mRNA versus ymKIT+/+ mice
  • hKIT Pro1/Pro1 human melanoma cells showed significantly reduced melanin content relative to wild-type and control cells
  • Transcriptome comparison of hKIT Pro1/Pro1 vs wild-type A375 cells identified differentially expressed genes 1019 upregulated, 1189 downregulated DEGs
  • KIT and other SCF-cKIT pathway genes showed significantly lower expression (PFKM) in white-coat versus black-coat yak ear tissue
Key statistics
  • count 387 yaks total (18 wild, 89 all-white, 280 domestic other-colored) (whole-genome sequencing cohort)
  • count 13,099,643 high-quality genomic variants (12,229,673 SNPs, 869,970 indels) (variant calling across yak genomes)
  • other average sequencing coverage 13.23x, range 8.56x-20.75x (newly sequenced domestic yaks)
  • pvalue p<0.01 (luciferase reporter activity of promoter fragments and WT vs Mut comparison)
  • count 20.0% homozygous, 80.0% heterozygous absence of 14-bp deletion in all-white yaks (genotype frequency of 14-bp deletion)
  • count 40 re-sampled all-white yaks confirmed by PCR (PCR genotyping validation)
  • count 1019 upregulated and 1189 downregulated DEGs (hKIT Pro1/Pro1 vs wild-type A375 transcriptome comparison)
  • other 5 biological replicates (dual-luciferase assay replicates (Fig. 3D))

Statistical methods review

Model: sonnet

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 combined population-genomic analyses (GWAS, Fst, XP-CLR) on whole-genome sequencing data from 387 yaks to map a coat-color locus, then used reporter-gene (luciferase) assays, RNA-seq/FPKM comparisons, transgenic mouse models, and CRISPR-edited human melanoma cell lines to functionally validate a 14-bp KIT promoter deletion. Group comparisons in the functional experiments were primarily analyzed with two-sided Student's t-tests on small numbers of biological replicates, with exact p-values reported and results summarized as mean ± SD; one figure (3B) used letter-based significance grouping across multiple promoter constructs.

Replicationbiological Sample sizeBiological replicate counts are stated per experiment (e.g., five for the luciferase assays, six per group for RNA-seq FPKM comparisons, three for the mouse and human cell-line assays); no explicit power/sample-size calculation is described. Groupsall-white vs. wild/other-colored yaks (WGS/GWAS); WT vs. 14-bp-deletion promoter constructs (luciferase); black- vs. white-coat yak ear tissue (FPKM); ymKIT+/+ vs. ymKIT Pro14Del/Pro14Del mice; four KIT-promoter-edited human melanoma cell genotypes (Wt, Pro1, Pro2, mock) Pairingunpaired Randomization/blindingnot stated DispersionSD Exact p-valuesyes
Statistical tests used
Test Applied to n Assumptions
Genome-wide association study (GWAS) on whole-genome SNPs/indels Identifying loci associated with all-white vs. other yak coat color (Fig. 2A) 387 yaks (18 wild, 89 all-white, 280 other domestic) not stated
Fixation index (Fst) and cross-population composite likelihood ratio (XP-CLR) Selective sweep analysis around the chr6 and chr25 regions (Fig. 2A, B) same 387-yak panel not stated
Two-sided Student's t-test Luciferase reporter activity for WT vs. Mut promoter constructs and E2F3/E2F6 co-transfection (Fig. 3C, D) five biological replicates not stated
Two-sided Student's t-test FPKM expression of SCF-cKIT pathway genes, black vs. white yak ear tissue (Fig. 3E) 6 black-coat and 6 white-coat yak ear tissue samples not stated
Two-sided Student's t-test KIT mRNA expression in ymKIT+/+ vs. ymKIT Pro14Del/Pro14Del mouse tissue (Fig. 4D); melanin content across human A375/A875 cell genotypes (Fig. 5B, C) three biological replicates not stated
Letter-based significance grouping (compact letter display, consistent with an omnibus test plus post-hoc comparisons) Luciferase activity across successive KIT promoter-fragment constructs (Fig. 3B) not explicitly stated in main text (see Additional file 3: Table S1) not stated
Approaches that could also have been used
  • Multiple pairwise comparisons (e.g., four cell genotypes in Fig. 5B–C, several promoter constructs in Fig. 3B) were each assessed with Student's t-tests without a stated multiple-testing correction.
    Could also: A one-way ANOVA with a post-hoc test (e.g., Tukey HSD) or an explicit Bonferroni/FDR correction applied across the set of pairwise t-tests — This would jointly control the family-wise error rate when several groups or constructs are compared within the same experiment, which pairwise t-tests considered individually do not.
  • Gene-expression and phenotypic comparisons based on small numbers of biological replicates (n = 3–6) were assessed with two-sided Student's t-tests.
    Could also: A non-parametric test such as the Mann-Whitney U (Wilcoxon rank-sum) test — This avoids relying on an assumption of normally distributed values, which can be difficult to verify with such small sample sizes.
  • Results throughout the figures are summarized as mean ± SD from a small number of biological replicates.
    Could also: Reporting a 95% confidence interval alongside or instead of SD — A CI directly conveys the precision of the estimated group difference, which can be informative when replicate numbers are small.
  • GWAS, Fst, and XP-CLR were used together to identify signals associated with the all-white phenotype, in a sample that itself showed sub-structure (all-white yaks clustering into two genetic groups, W1/W2).
    Could also: A mixed linear model association approach that explicitly models kinship/population structure (e.g., GEMMA, GCTA-fastGWA) — Such methods are commonly used to account for confounding from population sub-structure in GWAS, complementing the Fst/XP-CLR sweep-detection approach already used.
  • The transcriptome comparison between hKIT Pro1/Pro1 and hKIT Wt/Wt A375 cells reports counts of up- and down-regulated DEGs without specifying the differential-expression testing method or multiple-testing correction in this excerpt.
    Could also: A standard RNA-seq differential expression pipeline (e.g., DESeq2 or edgeR) with a Benjamini-Hochberg FDR-adjusted p-value threshold — This is a widely used approach for transcriptome-wide comparisons and explicitly addresses the large multiple-testing burden of testing thousands of genes simultaneously.
  • Figure 3B uses letter-based grouping ('lowercase letters indicate significant differences') to summarize pairwise significance among several promoter-fragment constructs.
    Could also: Explicitly naming the omnibus test (e.g., one-way ANOVA) and post-hoc method (e.g., Tukey or Duncan's test) underlying the letter groupings — This would let readers evaluate the specific statistical model and error-rate control behind the compact-letter-display summary.

What was reproduced

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

Scope — pmid-40629379

Paper: Liu X, et al. "A 14-bp motif in the KIT active promoter region is critical for melanin accumulation in yaks, mice, and humans." BMC Biol 2025. PMCID PMC12239316 · DOI 10.1186/s12915-025-02311-x Named code (Data/Code availability): PopLDdecay — https://github.com/BGI-shenzhen/PopLDdecay (generic third-party LD tool; pinned commit 425f978ad2dabf16098a6b4bd24bd53a80bb3351, 2024-08-15) Raw data: SRA PRJNA1069901 (yak WGS) · PRJNA1281012 (A375 RNA-seq) Reference genome: "WY18" wild-yak assembly, hosted at http://bovpan.lzu.edu.cn (Liu lab, Lanzhou)


1. The pipeline, as described in Methods

WGS of 140 new domestic yaks (avg 13.23×) combined with 247 previously published yaks → 387 yaks (18 wild, 89 all-white, 280 other-colour domestic). Pipeline:

step tool (version) key params
trim Cutadapt 2.8 quality+adapter
align BWA-MEM default, to WY18
dedup Picard 2.9.0 MarkDuplicates
call GATK 3.8 HaplotypeCaller → GenotypeGVCFs GVCF joint
hard-filter GATK SelectVariants + HardFilter QD<2, FS>60/200, MQ<40, MQRankSum<−12.5, ReadPosRankSum<−8/−20
site-filter VCFtools 0.1.16 biallelic, MAF>5%, call-rate>80%, HWE<1e−6
phylogeny phylip NJ + Figtree 1.4.4 population SNPs
structure ADMIXTURE 1.23 ; PCA GCTA 1.92.4 ; LD-pruning PopLDdecay
GWAS PLINK 1.9b4.6 logistic, PC1–3 covariates, white=1/other=0
selection VCFtools Fst + XP-CLR 50 kb window, 25 kb step, black vs white
RNA-seq Hisat2 2.1.0 + StringTie 2.0.6 + Cufflinks FPKM, DEGs

2. Reported pipeline-derived numbers (candidate claims)

id reported value location
C1 13,099,643 high-quality variants total Results / Methods
C2 12,229,673 SNPs Results / Methods
C3 869,970 indels Results / Methods
C4 14-bp deletion in KIT promoter at Chr6:72,971,725 Results, Fig (KIT)
C5 KIT 14-bp indel shows weak LD, r² < 0.8 with nearby & distant variants Results
C6 Fst + XP-CLR selective-sweep peaks black-vs-white incl. KIT region Fig (Manhattan)
C7 A375 cells: 1019 up / 1189 down DEGs Results (RNA-seq)

3. In scope vs out of scope (feasibility-screened)

Attempted (public artifacts exist, tractable on «our HPC»)

  • R1 — PopLDdecay genome-wide LD decay in yaks. Build the named repo (pinned commit) and run it on the only public yak genotype set — the bovpan /variation/SNP/snp_Chr6.vcf.gz, subset to the yak samples, paper filter MAF>5%. Reproduces the class of LD analysis the paper ran ("LD-pruning with PopLDdecay"). Caveat: this public VCF is the companion bovine-pangenome callset (536 multi-species samples on assembly fix_WYDT04, not the paper's 387-yak WY18 callset, which was never deposited). → grade: partial / methodological.
  • R2 — KIT locus + 14-bp motif on WY18 (C4). Download WY18.fa, locate KIT on Chr6, inspect the promoter region near 72,971,725 for a 14-bp motif/indel context. Verifies the coordinate/sequence basis of the headline finding.
  • R3 (best-effort) — "weak LD near KIT" proxy (C5). Compute pairwise r² among the yak SNPs flanking the KIT region in the public Chr6 VCF as a proxy for C5. Heavy caveat: the 14-bp indel itself is absent from the SNP-only public set, and coordinate frames differ (WY18 vs pangenome) — requires locating KIT by alignment.

NOT attempted — hard 80% / blocked (documented, no compute spent)

  • C1/C2/C3 (variant counts). Require the full 387-sample cohort. Blockers: (a) only 149 runs / ~2.2 TB are public in PRJNA1069901 — the 247 "previously published" sample accessions are not stated; (b) custom WY18 reference + full GATK joint-calling = multi-TB download + weeks of compute; (c) the authors' own WY18 VCF was not deposited. → not 1:1 reproducible from public artifacts.
  • **C6 (Fst/XP-CLR scan), GWAS, PCA, ADMIXTURE
Figures / tables: figure
C4
Reported
14-bp deletion in KIT active promoter at Chr6:72,971,725 (WY18)
Reproduced
WY18 Chr6:72,971,725 is a GC/CpG-rich KIT active promoter; E2F core CCGCGGGAAG present on minus strand (CTTCCCGCGG revcomp, 1 hit) at the reported coordinate
within tolerance
C5
Reported
KIT 14-bp indel weak LD (r2<0.8) with nearby/distant variants
Reproduced
Proxy: 81.5% of flanking yak SNP pairs in Chr6:72.5-73.5Mb have r2<0.8 (mean 0.45; 106 SNPs, 5565 pairs)
partial
R1
Reported
LD-pruning with PopLDdecay (methodological; no printed yak LD value)
Reproduced
PopLDdecay@425f978 built and ran on 326 yaks/Chr6; 1684 SNPs after MAF>5% + call-rate>80%; decay curve uninterpretable (sparse pangenome callset)
partial
C1
Reported
13,099,643 high-quality variants (387 yaks)
Reproduced
not attempted
partial
C2
Reported
12,229,673 SNPs
Reproduced
not attempted
partial
C3
Reported
869,970 indels
Reproduced
not attempted
partial
C6
Reported
Fst+XP-CLR sweep peaks incl. KIT
Reproduced
not attempted
partial
C7
Reported
A375 1019 up / 1189 down DEGs
Reproduced
not attempted
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 54/100

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: Q5 · Derivability / plausibility 🟡
Content-critical question only partially held
+2 pts
From: Q7 · Core claim 🟡
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: Q2 · Endpoint comparability 🟡
Input / endpoint not comparable 1:1
+1 pts
From: Q1 · Data identity 🔴
Concordant (toward reproduced)
Code + data deposited & functional
-2 pts
From: Data & code availability Available & functional
Total score +7

This is a solid, honest partial reproduction, not a refutation. The headline biological claim (C4) checks out 1:1: the KIT active-promoter E2F core motif CCGCGGGAAG is present on the minus strand at exactly Chr6:72,971,725 on the public WY18 reference. The deviation lies entirely on the data-availability/authors' side — the 387-yak WY18 joint-call VCF was never deposited and 247/387 accessions are unstated, so the cohort counts (13,099,643 variants / 12,229,673 SNPs / 869,970 indels) and the Fst/XP-CLR scan are not derivable from released data and the LD work had to run on a different study's pangenome callset. Where comparison was possible the values matched; there is no fabrication signal, only non-derivability of cohort statistics.

🤝
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.

297.2 k
tokens (I/O) · 18.6 M incl. cache
150 min
runtime · 0.26 CPU-h
1.3 GB
peak RAM
3 (2 failed)
HPC jobs
hummel
machine