Tabula Sapiens reveals the non-coding RNA landscape across 22 human organs and tissues

Name: Verified Rigor Record — Tabula Sapiens reveals the non-coding RNA landscape across 22 human organs and tissues
Item: Tabula Sapiens reveals the non-coding RNA landscape across 22 human organs and tissues
Rating: 75
Author: Αlpha¹

Lee, J.; Mantri, M.; Murthy, K.; Seeker, L. A.; Crowley, G.; Jones, R. C.; Tabula Sapiens Consortium, ; Quake, S. R.

Paper source

Read PDF

Αlpha¹ Rigor Review

Rigor Review Verified

Critical

High-priority

Action items to resolve · Critical / High

Confidence: moderateEngine: 7.6.0Source: full text

Not verified in this report

References were not verified against Crossref/OpenAlex.
No reported statistical tests were found to recompute.
No data or code availability links were detected to verify.

This Alpha1 Rigor Review uses AI Rigor Reviewers trained on a curated corpus of high-fidelity and retracted papers, with expert supervision and curation. It can still make mistakes; verify each finding against the source before relying on it.

The paper presents a comprehensive single-cell atlas of non-coding RNA expression across 22 human tissues from a single donor, with a clear scientific premise and ethical approvals. However, the absence of a data availability statement and repository deposit is a critical gap for a resource paper, and reporting of biological variables and statistical analysis is incomplete.

This assessment synthesizes three independent reviewer evaluations. All reviewers agreed on most dimensions, but diverged on study design, key resources, data code availability, and reporting transparency. The synthesis weighed the specific evidence from each reviewer to reach a consensus that is conservative where evidence was ambiguous.

10 major claims checked against the paper's own evidence: all adequately supported.

supportedAbstractReviewers 1, 3
The dataset enables joint analysis of protein-coding and non-coding transcriptomes at single-cell and subcellular resolution.
The paper demonstrates through analyses of cell type specificity, nuclear compartmentalization, and tRNA repertoires that the data indeed captures both coding and non-coding transcriptomes.
Evidence: Results sections on cell type specificity, nuclear compartmentalization, tRNA analysis, and cell cycle/senescence show joint analysis.
“By simultaneously profiling both polyadenylated and non-polyadenylated transcripts, the resulting dataset enables joint analysis of the protein-coding and non-coding transcriptomes at single-cell and subcellular resolution.”
supportedResults, Cell type specificity of ncRNAsReviewers 1, 2, 3
A greater proportion of non-coding genes are differentially expressed by single cell types compared to protein-coding genes.
The paper shows that when normalized by the number of annotated genes, a larger fraction of lncRNAs than protein-coding genes are unique DEGs, and non-coding genes comprise 75.6% of unique DEGs.
Evidence: Figure 3 and associated text: 'When normalized by the total number of annotated genes of each biotype in the human genome, a larger fraction of lncRNAs than protein-coding genes were DEGs uniquely in a single cell type.'
“When normalized by the total number of annotated genes of each biotype in the human genome, a larger fraction of lncRNAs than protein-coding genes were DEGs uniquely in a single cell type (“unique DEGs”).”
supportedResults, Cell type specificity of tRNA repertoiresReviewers 1, 2, 3
tRNA repertoires are cell type-specific and this specificity is not simply explained by differences in codon usage across cell types.
The paper demonstrates cell type-specific tRNA profiles via correlation analysis and shows that theoretical translation efficiency is not significantly higher than random pairings, indicating other factors drive specificity.
Evidence: Results section on tRNA cell type specificity and Figure 5: 'tTEs calculated from our dataset indicated well-matched tRNA supply and demand for most cell types... observed tTEs were not elevated relative to values calculated from random pairings of tRNA repertoires and amino acid usages.'
“Taken together, these analyses suggest that while tRNA pools are globally well-aligned to translational demands, other factors besides translational demands shape tRNA repertoires in a cell type-specific manner.”
supportedResults, ncRNAs dynamics throughout the cell cycle and in cells expressing senescence-associated transcriptsReviewer 1
We identified non-coding genes with putative roles in cell division and growth arrest.
The paper identifies ncRNAs consistently upregulated in S phase and in CDKN2A+ MKI67- cells, providing candidates for further study.
Evidence: Results section on cell cycle and senescence: 'multiple ncRNA genes that were consistently upregulated in each phase... a total of 269 putative senescence-associated ncRNA genes enriched in CDKN2A+ MKI67− cells.'
“We identified multiple ncRNA genes that were consistently upregulated in each phase ()... a total of 269 putative senescence-associated ncRNA genes enriched in CDKN2A+ MKI67− cells across cell types ().”
supportedDiscussionReviewer 1
Our work establishes a resource for investigating the landscape of non-coding RNAs across a diverse set of human tissues and cell types.
The paper presents a comprehensive dataset and multiple analyses that demonstrate its utility, establishing it as a resource for the community.
Evidence: The entire paper, including the dataset generation and the vignettes showing utility.
“Our work establishes a landmark resource for surveying the human non-coding RNA landscape.”
supportedAbstractReviewer 2
The dataset enables joint analysis of the protein-coding and non-coding transcriptomes at single-cell and subcellular resolution.
The paper presents evidence of profiling both polyadenylated and non-polyadenylated transcripts from single cells and nuclei, with detailed methods and quality control.
Evidence: The paper describes the TotalX protocol and single-nucleus sequencing, and presents data showing detection of both coding and non-coding RNAs.
“By simultaneously profiling both polyadenylated and non-polyadenylated transcripts, the resulting dataset enables joint analysis of the protein-coding and non-coding transcriptomes at single-cell and subcellular resolution.”
supportedAbstractReviewer 2
The work establishes a resource for investigating the landscape of non-coding RNAs across a diverse set of human tissues and cell types.
The paper provides a comprehensive dataset and demonstrates its utility through multiple analyses, though the single-donor nature limits diversity.
Evidence: The paper describes the dataset composition and presents several analyses demonstrating its utility.
“Our work establishes a resource for investigating the landscape of non-coding RNAs across a diverse set of human tissues and cell types.”
supportedAbstractReviewer 2
Nuclear compartmentalization of ncRNAs is cell type-dependent.
The paper provides evidence from comparing single-cell and single-nucleus data, showing cell type-specific patterns of nuclear enrichment for many ncRNAs.
Evidence: Figures 4 and S10 show cell type-specific nuclear enrichment patterns and cosine similarity analysis.
“we compared single-cell and single-nucleus data from the same samples to infer subcellular localization patterns, revealing cell type-dependent nuclear and cytoplasmic enrichment of specific non-coding RNAs.”
supportedAbstractReviewer 2
Non-coding genes with putative roles in cell division and growth arrest are identified.
The paper identifies ncRNAs consistently upregulated in specific cell-cycle phases and in CDKN2A+ MKI67- cells, though functional validation is not provided.
Evidence: Figures 7 and 8 show cell-cycle phase-specific and senescence-associated ncRNAs.
“we characterized dynamic expression patterns of non-coding RNAs across the cell cycle and senescence-associated cell states, identifying non-coding genes with putative roles in cell division and growth arrest.”
supportedAbstract, final sentenceReviewer 3
Non-coding RNAs with putative roles in cell division and growth arrest are identified.
The paper identifies specific lncRNAs consistently upregulated in S phase across cell types and ncRNA genes enriched in senescence-associated cells.
Evidence: Figure 6 and results describing cell-cycle and senescence analysis.
“identified non-coding genes with putative roles in cell division and growth arrest.”

Rigor

Eight Dimensions3 pass · 4 warn · 1 fail

10 major claims10 supported

6 copyedit issues flagged: mostly consistency, clarity.

Alpha¹ Rigor Action ItemsFinal pass before journal submission

The published manuscript is a valuable resource but has a critical gap: no data availability statement or repository accession numbers are provided, which limits reproducibility and reuse. An informed reader should weigh this omission heavily; the authors should issue a correction or add a data availability statement with accession numbers to the published version. Other issues (incomplete biological variable reporting, statistical reporting) are less severe but should be noted.

1.
HIGHdata code
Provide a data availability statement with GEO accession numbers for raw sequencing data and a repository (e.g., Zenodo) for processed count matrices, and deposit the code in a public repository (e.g., GitHub) with a persistent identifier.
This is a critical omission for a resource paper; without it, the data cannot be verified or reused, which undermines the paper's primary contribution.
2.
HIGHrigor
Add a scientific justification for using a single male donor in the Results or Discussion, or explicitly acknowledge the lack of justification as a limitation.
The use of a single sex without justification limits the generalizability of the findings and is a recognized reporting gap.
3.
HIGHstatistics
Report exact p-values (e.g., p = 0.003) instead of only thresholds, and provide confidence intervals for effect sizes (e.g., log2 fold change) in key comparisons.
Imprecise statistical reporting prevents readers from fully assessing the evidence and is a common concern in peer review.
4.
HIGHreporting
Include a statement verifying that test assumptions (e.g., normality, equal distributions for Wilcoxon) were checked, or justify the use of non-parametric tests.
Failure to verify assumptions can lead to incorrect inferences, even with non-parametric tests.
5.
MEDIUMreporting
Define error bars in all figures (e.g., SD, SEM, CI) and state the rationale for the chosen measure.
Undefined error bars make figures uninterpretable and reduce the transparency of data presentation.
6.
MEDIUMreporting
Add a conflict of interest statement (even if 'none declared') to the manuscript.
A missing COI statement is a minor but easily fixable reporting gap that journals require.
7.
MEDIUMreporting
Reference a relevant reporting guideline (e.g., STROBE for observational studies) in the methods section.
While not mandatory, referencing a reporting guideline enhances transparency and reproducibility.
8.
MEDIUMrigor
Report the donor's weight and any additional health status details if available, or note that they were not recorded.
Complete demographic information is important for assessing the context of the biological sample.
9.
MEDIUMreporting
Clarify the distinction between biological and technical replicates in the study design, if applicable.
Clarity on replicates helps readers understand the independence of observations.
10.
LOWcopyedit
Remove the phrase 'pmc 1 preprint 1' from the title.
This appears to be a formatting artifact from the preprint server that should not appear in the final manuscript.
11.
LOWcopyedit
Add a comma after 'RN7SL' in the sentence listing highly abundant ncRNA species.
Minor punctuation consistency improves readability.
12.
LOWcopyedit
Add a comma after 'Stanford University' in the ethics approval sentence for clarity.
Improves grammatical clarity without changing meaning.

Full text (54,333 chars)Analyzed: Jul 7, 2026

Αlpha¹ assesses methodological rigor, not the importance of the findings. See how we calibrate →