CoChat
Sign Up
Cited by 151MicroRNA in disease regulationPlant Molecular Biology ResearchCancer-related molecular mechanisms research

Toward the human cellular microRNAome

Open access
McCall, Matthew N., Kim, Min-Sik, Adil, Mohammed, Patil, Arun H., Lu, Yin, Mitchell, Christopher J., Leal-Rojas, Pamela, Xu, Jinchong +9 more·2017·Genome Research
DOI 10.1101/gr.222067.117PMC PMC5630040

Published version of a preprint

Open this paper in CoChat — free accountOpen PDF

Abstract

MicroRNAs are short RNAs that serve as regulators of gene expression and are essential components of normal development as well as modulators of disease. MicroRNAs generally act cell-autonomously, and thus their localization to specific cell types is needed to guide our understanding of microRNA activity. Current tissue-level data have caused considerable confusion, and comprehensive cell-level data do not yet exist. Here, we establish the landscape of human cell-specific microRNA expression. This project evaluated 8 billion small RNA-seq reads from 46 primary cell types, 42 cancer or immortalized cell lines, and 26 tissues. It identified both specific and ubiquitous patterns of expression that strongly correlate with adjacent superenhancer activity. Analysis of unaligned RNA reads uncovered 207 unknown minor strand (passenger) microRNAs of known microRNA loci and 495 novel putative microRNA loci. Although cancer cell lines generally recapitulated the expression patterns of matched primary cells, their isomiR sequence families exhibited increased disorder, suggesting DROSHA- and DICER1-dependent microRNA processing variability. Cell-specific patterns of microRNA expression were used to de-convolute variable cellular composition of colon and adipose tissue samples, highlighting one use of these cell-specific microRNA expression data. Characterization of cellular microRNA expression across a wide variety of cell types provides a new understanding of this critical regulatory RNA species.

AI snapshot

Written from the paper’s full text.

TL;DR

This study establishes a human cell-specific microRNA expression landscape by analyzing small RNA-seq across primary cells, cancer/immortalized cell lines, and tissues. It identifies cell-restricted and ubiquitous microRNAs, links expression to superenhancers, reports 495 high-confidence putative novel microRNAs, and shows substantial isomiR diversity that differs between primary and cancer cells.

Key findings

  • Across 46 primary cell types, 42 cancer/immortalized cell lines, and 26 tissues, 2319 of 2546 known microRNAs (miRBase v21) were detected at ≥1 read per million microRNA reads in at least one sample; 320 microRNAs reached ≥1000 RPM in at least one of 46 normal cell types, and 94 (29%) were present in only a single cell class. Six microRNAs were present in all 46 cell types at ≥1000 RPM: miR-107, miR-103a-3p, miR-103b, miR-191-5p, miR-21-5p, and miR-92a-3p.
  • MicroRNA expression strongly correlated with adjacent superenhancer activity: expression was significantly higher when a superenhancer was within 40 kb of the microRNA locus (Wilcoxon rank-sum test, P < 2.2−16), and cell-type-restricted microRNAs matched cell-specific superenhancer activity.
  • Novel microRNA discovery from 1.2 billion unmapped reads yielded 21,338 miRDeep2 loci, which were refined to 495 highest-confidence putative novel microRNAs; 207 unassigned passenger microRNAs from known loci were also found. Validation included miRanalyzer overlap, Argonaute CLIP-seq evidence, PhyloP conservation, DICER1-knockout dependence, and PCR of nine novel microRNAs.
  • IsomiR analysis of 126 primary cell and 82 cancer/immortalized cell samples showed that the miRBase v21 canonical sequence was the most abundant isomiR in only 33% of primary-cell microRNAs, and 37% never had the canonical sequence as most abundant; cancer/immortalized cells had a ∼20% increase in disorder over primary cells (P < 0.005, Wilcoxon rank-sum test).

Methods

  • The study combined new small RNA-seq from 39 primary cell samples with public Sequence Read Archive/Array Express data, processing 694 samples through miRge and retaining 530 for analysis: 161 primary cell datasets (46 main cell types), 100 cancer/immortalized cell line datasets (42 cell lines), and 269 tissue datasets (26 tissues).
  • MicroRNA expression states were discretized with DEXUS; batch effects were addressed with RUV normalization and SVA; clustering used t-SNE. Cell-specific signatures were used with CIBERSORT to de-convolute adipose and colon tissue composition.
  • Novel microRNA detection used miRDeep2 and miRanalyzer on unmapped reads, followed by novoMiRank, PhyloP conservation, Argonaute CLIP-seq, DICER1-knockout data, and PCR validation. IsomiR families were analyzed for templated length variants, 5′ start variation, and nontemplated 3′ nucleotide additions.

Limitations

  • The paper states that much data came from public sources across different platforms and sequencing methods, creating technical variation; extensive normalization was used, but technical factors still drive some variation and clustering. Some cell types had few to no replicates, it is unknown how well cultured cells match in vivo cells, and many cell types (hepatocytes, neutrophils, pneumocytes) were unavailable.
  • The paper states that the 495 putative novel microRNAs are not definitive, many are likely false positives, and in vivo proof was beyond scope; requiring multiple evidence sources may also have lost true positives.
  • Evident from the text: superenhancer correlations are not proof of superenhancer activity on adjacent microRNAs, and the authors caution that the activity of these superenhancers on adjacent genes and microRNAs is generally unknown.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1101/gr.222067.117

Cites

60 of 60 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Claims the paper makes

2 of 47 cited sources checked · 1 supported · 0 flagged · accuracy not yet measured

  • Findingown data

    Among primary cell types, microRNA expression patterns generated four major groups: hematologic, mesenchymal, neural/ESC, and epithelial.

    “Akin to the DEXUS results, among primary cell types, microRNA expression patterns generated four major groups: hematologic, mesenchymal, neural/ESC, and epithelial.”
    Quote in the paperSubstantial claim
  • Findingown data

    Seven microRNAs (let-7a-5p, let-7c-5p, let-7b-5p, let-7f-5p, let-7g-5p, miR-26a-5p, and miR-30d-5p) were found in all tissues.

    “Seven of these microRNAs (let-7a-5p, let-7c-5p, let-7b-5p, let-7f-5p, let-7g-5p, miR-26a-5p, and miR-30d-5p) were found in all tissues.”
    Quote in the paperNot a substantial claim
  • Findingown data

    miR-126-3p, abundant only in endothelial cells and platelets, was present in 21 tissues.

    “Likewise, miR-126-3p, abundant only in endothelial cells and platelets, was present in 21 tissues, and miR-150-5p, abundant in lymphocytes, was present in 11 tissues.”
    Quote in the paperNot a substantial claim
  • Findingown data

    DEXUS method clustered cell types into hematologic, neural/embryonic stem cell, epithelial, and mesenchymal groups.

    “This method clustered cell types into hematologic, neural/embryonic stem cell (ESC), epithelial, and mesenchymal groups, identifying general patterns of microRNA expression.”
    Quote in the paperSubstantial claim
  • Findingown data

    There were 320 microRNAs that had an RPM ≥1000 in any of the 46 normal cell types.

    “There were 320 microRNAs that had an RPM ≥1000 in any of the 46 normal cell types.”
    Quote in the paperSubstantial claim
  • Classified as probably not a substantial claim. 19
  • May combine two claims. 32

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    MicroRNAs control essential cell processes in health and disease.

    “In this role, they control essential cell processes in health and disease (Ambros 2004; Mendell and Olson 2012).”
    Quote in the paperNot a substantial claim1 of 2 sources checked
    • cites Ambros 2004 source
    • cites Mendell and Olson 2012 source

  • May combine two claims.

    Interpretationcitation

    This number of new microRNAs seems impossibly high and likely mostly false positives and/or nonfunctional transcripts.

    “This number of new microRNAs seems impossibly high and likely mostly false positives and/or nonfunctional transcripts (Mullokandov et al. 2012).”
    Quote in the paperSubstantial claim1 of 1 sources checked
    • cites Mullokandov et al. 2012 source

      The Discussion (page unavailable in our copy) reports that almost 60% of miRNAs detected in the cells studied had no discernible suppressive activity, supporting concern that many detected miRNAs may have little biological activity. It does not assess the number of *new* miRNAs or conclude that most are false-positive identifications. The Results’ reference to possible false positives concerns a small set of sensor results, not the miRNA transcripts themselves. (LLM, conf=0.87)

      Experimental — accuracy not yet measured

  • May combine two claims.

    Findingown data

    miR-451a was abundantly present in 20 of 26 tissues but is from only one cell class (red blood cells).

    “miR-451a was abundantly present in 20 of 26 tissues but is from only one cell class (red blood cells).”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    Strong clustering by biological group was observed for all samples, overcoming most technical concerns.

    “Strong clustering by biological group was observed for all samples, overcoming most technical concerns.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    miR-150-5p, abundant in lymphocytes, was present in 11 tissues.

    “and miR-150-5p, abundant in lymphocytes, was present in 11 tissues.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    The average z-score was right-shifted to 1.07 for a subset of the 21,338 loci, while for the 495 highest confidence loci it was 0.90.

    “We found the average z-score was right-shifted to 1.07 (indicating less similarity) for a subset of the 21,338 loci, while for the 495 highest confidence loci it was 0.90, indicating more consistency with miRBase microRNAs (Supplemental Fig. S8A,B; Backes et al. 2016).”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Backes et al. 2016

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • May combine two claims.

    Findingcitation

    miR-199b-3p (+1 shift), miR-181c-3p (+1 shift), and miR-302a-5p (+3 shift) had highly abundant reads containing a completely different seed sequence than the one currently assigned.

    “This included miR-199b-3p (+1 shift), miR-181c-3p (+1 shift), and miR-302a-5p (+3 shift), all of which had highly abundant reads containing a completely different seed sequence than the one currently assigned, with strong implications for the targeting of genes (Tan et al. 2014).”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Tan et al. 2014 source

  • May combine two claims.

    Findingown data

    Eight primary T lymphocyte samples and 14 T lymphocyte malignancy samples revealed moderate to strong correlation but with separate primary and cancer-derived cell clustering.

    “Eight primary T lymphocyte samples and 14 T lymphocyte malignancy samples also revealed moderate to strong correlation but with separate primary and cancer-derived cell clustering (Supplemental Fig. S6C).”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    Thirty-four putative novel microRNAs were also adjacent to a known microRNA.

    “Thirty-four were also adjacent to a known microRNA.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    These modifications could vary widely between both primary and cancer/immortalized cells as a reflection of either biological differences or technical factors.

    “but could vary widely between both primary and cancer/immortalized cells as a reflection of either biological differences or technical factors (Fig. 4C,D).”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    98% of sequencing runs were performed on an Illumina Genome Analyzer or HiSeq system and >74% of libraries were generated using an Illumina sequencing kit.

    “Certain aspects of the data also helped with normalization, including 98% of sequencing runs being performed on an Illumina Genome Analyzer or HiSeq system and >74% of libraries being generated using an Illumina sequencing kit.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Findingown data

    15 were orthologous to a different species' microRNA (primarily primate).

    “and 15 were orthologous to a different species’ microRNA (primarily primate) (Supplemental Tables S8, S9).”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    While miR-126 is highly conserved throughout chordates, JHU_ID_23828 is only conserved among primates.

    “Additionally, while miR-126 is highly conserved throughout chordates, JHU_ID_ 23828 is only conserved among primates (Fig. 3G).”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data

    A test for difference in proportions between wild-type and DICER1 knockout samples revealed the proportion in wild-type dropped from 0.151 to 0.047 in the DICER1 samples, representing a significant decrease (P < 2.2−16).

    “A test for the difference in proportions between wild-type and DICER1 knockout samples revealed the proportion in wild-type dropped from 0.151 to 0.047 in the DICER1 samples, representing a significant decrease (P < 2.2−16).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    PhyloP conservation data show that beyond the most ideal microRNAs in miRGeneDB, there is a lack of genetic conservation, with overlap among a large segment of miRBase microRNAs and the novel microRNAs reported here and in a prior large study.

    “The PhyloP conservation data (Fig. 3B) show that, beyond the most ideal microRNAs identified in miRGeneDB, there is a lack of genetic conservation, with overlap in a large segment of miRBase microRNAs and the novel microRNAs reported here and in a prior large study (Londin et al. 2015).”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Londin et al. 2015 source

      Checked before — its earlier verdict is held back until a new check runs.

  • May combine two claims.

    Findingcitation

    JHU_ID_23828-3p was validated by PCR and is located in intron 8 of the EGFL7 gene, approximately 900 bp from mir-126 and within the same pri-miRNA transcript.

    “As an example of potentially interesting novel microRNAs, JHU_ID_23828-3p was validated by PCR and is located in intron 8 of the EGFL7 gene, approximately 900 bp from mir-126 and within the same pri-miRNA transcript (Fig. 3D,E; Chang et al. 2015).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Chang et al. 2015 source

  • May combine two claims.

    Findingown data

    The most abundant isomiR was widely variable between cells and often incongruent with the expected sequence.

    “The most abundant isomiR was widely variable between cells and often incongruent with the expected sequence.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    This was quite variable and rarely >90%.

    “This was quite variable and rarely >90%, suggesting methods that fail to acquire isomiRs significantly underestimate the presence of microRNAs in a variable fashion (Mestdagh et al. 2014).”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Mestdagh et al. 2014 source

  • May combine two claims.

    Findingown data

    An analysis of 12 primary and three immortalized fibroblast cell line microRNA signatures identified overall strong correlation (RUV-corrected, log2-normalized, pairwise R > 0.80-0.99), with immortalized lines being slightly less correlated.

    “An analysis of 12 primary and three immortalized fibroblast cell line microRNA signatures identified overall strong correlation (RUV-corrected, log2-normalized, pairwise R > 0.80–0.99), with the immortalized lines being slightly less correlated (Supplemental Fig. S6A).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    There were more microRNAs that differed between primary and cancer-derived cells, including miR-150-5p which was 64-fold higher in primary T cells.

    “There were more microRNAs that differed between primary and cancer-derived cells (Supplemental Fig. S6D), including miR-150-5p which was 64-fold higher in the primary T cells, as has been reported (He et al. 2014).”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites He et al. 2014

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • May combine two claims.

    Findingown data

    Using validation steps, the number of candidate novel microRNAs was reduced to 495 highest-confidence miRNAs, roughly one new microRNA per sample evaluated.

    “Using a number of validation steps, we reduced this number down to 495 highest confidence miRNAs, which is roughly one new microRNA per sample evaluated.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingown data

    The remaining miRBase microRNAs had modest conservation and were similarly distributed to both the putative novel microRNAs reported here and the 3707 novel TJU microRNAs.

    “The remaining miRBase microRNAs had modest conservation and were similarly distributed to both the putative novel microRNAs reported here and the 3707 novel microRNAs reported as TJU microRNAs (Fig. 3B; Londin et al. 2015).”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Londin et al. 2015 source

      Checked before — its earlier verdict is held back until a new check runs.

  • May combine two claims.

    Findingown data

    The data demonstrate general consistencies, but not without some concerning differences, between primary cells and immortalized/cancer cell lines.

    “Our data also demonstrate general consistencies, but not without some concerning differences, between primary cells and immortalized/cancer cell lines.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    MicroRNAs are an established class of small regulatory RNAs that bind mRNAs within the RISC complex and repress protein production.

    “MicroRNAs are an established class of small regulatory RNAs that, within the RISC complex, bind mRNAs and repress protein production (Valencia-Sanchez et al. 2006).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Valencia-Sanchez et al. 2006 source

  • Probably not a substantial claim: may not pass S2 (checkable).

    Interpretationreasoning

    Because all tissues are composed of multiple, unique cell types, it is essential to understand from which cell the microRNA signal is obtained.

    “Because all tissues are composed of multiple, unique cell types, it is essential to understand from which cell the microRNA signal is obtained.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Premisecitation

    The most thorough microRNA localization projects performed small RNA-seq on over 250 libraries from 26 organ systems but sequenced only approximately 1200 reads per library.

    “The most thorough of these microRNA localization projects performed small RNA library sequencing (RNA-seq) on over 250 libraries from 26 organ systems. However, this nascent effort sequenced only ∼1200 reads per library (Landgraf et al. 2007).”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Landgraf et al. 2007 source

  • Classified as probably not a substantial claim.

    Premisecitation

    IsomiRs are related sequences with mostly 5′ and 3′ nucleotide modifications that collectively make up the totality of a given microRNA.

    “IsomiRs are related sequences with mostly 5′ and 3′ nucleotide modifications that collectively make up the totality of a given microRNA (Neilsen et al. 2012).”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Neilsen et al. 2012 source

  • May combine two claims.

    Premisereasoning

    MicroRNA expression has been predominantly characterized in tissues, with no comprehensive cellular studies.

    “MicroRNA expression has been predominantly characterized in tissues, with no comprehensive cellular studies.”
    Quote in the paperSubstantial claim0 of 3 sources checked
    • cites Lagos-Quintana et al. 2002

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

    • cites Barad et al. 2004

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

    • cites Liu et al. 2004 source

  • Classified as probably not a substantial claim.

    Premisecitation

    Superenhancers are dense genomic regions of transcription factor binding sites that have a multiplicative effect on increasing adjacent gene expression.

    “Superenhancers are dense genomic regions of transcription factor binding sites that have a multiplicative effect on increasing adjacent gene expression (Whyte et al. 2013; Suzuki et al. 2017).”
    Quote in the paperSubstantial claim0 of 2 sources checked
    • cites Whyte et al. 2013 source

      Not checked in the free run.

    • cites Suzuki et al. 2017 source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Due to inexact cutting of DICER1 and DROSHA and nucleotide additions/modifications, a collection of mostly similar sequences (with most diversity on the 3′ end) make up the isomiR family of a microRNA.

    “Due to the inexact cutting of DICER1 and DROSHA and nucleotide additions/modifications, a collection of mostly similar sequences (with most diversity on the 3′ end) make up the isomiR family of a microRNA (Morin et al. 2008; Neilsen et al. 2012).”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites Morin et al. 2008 source
    • cites Neilsen et al. 2012 source

  • Classified as probably not a substantial claim. May combine two claims.

    Premisecitation

    DICER1 knockouts are known to significantly repress 3p driver microRNA levels while allowing for longer 5p microRNAs that are loaded into Ago and trimmed back from the 3p end.

    “DICER1 knockouts are known to significantly repress 3p driver microRNA levels while allowing for longer 5p microRNAs that are loaded into Ago and trimmed back from the 3p end (Hill et al. 2009).”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites Hill et al. 2009 source

  • May combine two claims.

    Study design and conductown data

    The data matched ground truths about specific microRNA expression in certain cell types, including miR-192 in epithelial cells, miR-126 in endothelial cells and platelets, and miR-9 in brain.

    “Our data matched certain “ground truths” about specific microRNA expression in certain cell types, including miR-192 in epithelial cells, miR-126 in endothelial cells and platelets, and miR-9 in brain (McCall et al. 2011a; Haider et al. 2014; Kent et al. 2014).”
    Quote in the paperSubstantial claim0 of 3 sources checked
    • cites McCall et al. 2011a

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

    • cites Haider et al. 2014 source
    • cites Kent et al. 2014 source

  • Probably not a substantial claim: may not pass S2 (checkable), S3 (load-bearing).

    Interpretationown data

    This technical driver of isomiR type will challenge those groups working toward a standard isomiR nomenclature.

    “This technical driver of isomiR type will challenge those groups working toward a standard isomiR nomenclature.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Interpretationown data

    The lack of conservation among many miRBase and novel microRNAs suggests either that much of miRBase is inaccurate or that conservation is not required for microRNA authenticity.

    “Either this suggests that much of miRBase is inaccurate or that, while conservation provides evidence for a microRNA to be authentic, it is not a requirement.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationown data

    These data demonstrate the importance and feasibility of solving for the cellular content of tissues to better understand the composition of the analyzed tissue.

    “Altogether, these data demonstrate the importance and feasibility of solving for the cellular content of tissues to better understand the composition of the analyzed tissue.”
    Quote in the paperSubstantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Premisecitation

    Another recent paper found relative quantification of microRNA RNA-seq across multiple laboratories to be remarkably accurate and reproducible, consistent with this work.

    “As well, another recent paper investigating microRNA RNA-seq across multiple laboratories found relative quantification to be “remarkably accurate and reproducible” (Giraldez et al. 2017), consistent with our work.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites Giraldez et al. 2017 source

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    The cell line H1264, a lung carcinoma cell line, has been reported as cross-contaminated with H157, a separate human lung carcinoma cell line.

    “The cell line H1264, which is a lung carcinoma cell line, has been reported as being cross-contaminated with H157, which is a separate human lung carcinoma cell line (ICLAC.org).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites ICLAC.org

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductown data

    A consensus adapter sequence could not be determined for 45 samples, and these samples were trimmed to 21 bp and excluded from isomiR analyses.

    “A consensus adapter sequence could not be determined for 45 samples, and the sequences were trimmed to 21 bp using the CutAdapt –u command (e.g., $ cutadapt <FILE>.fastq -u -14 -o <FILE>_cut.fastq for a 35-bp read length). These samples were excluded from isomiR analyses.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Study design and conductown data

    For microRNAs designated .5p/.3p, there were no passenger reads to distinguish the correct structure, so either location remains plausible.

    “The microRNAs designated as .5p/.3p were named as such, as each of these microRNA loci was identified in more than one sample, with two different pre-miRNA structures designating the sequence to the 5p or 3p arm. There were no “passenger” reads to distinguish the correct structure, so either location remains plausible.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Study design and conductown data

    The five estimated latent factors from RUV captured and adjusted for batch effects, as verified by examining biological clusters comprised of multiple experiments; RUV-normalized data clustered by biological cluster and not experimental batch.

    “We verified the ability of the five estimated latent factors to capture and adjust for batch effects by examining biological clusters comprised of multiple experiments. The RUV-normalized data clustered by biological cluster and not experimental batch.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationnone given

    The microRNA expression patterns from 42 cell types are the first step toward a complete understanding of microRNA expression across all cells and establishing a human cell atlas.

    “These microRNA expression patterns from 42 cell types are the first step toward a complete understanding of microRNA expression across all cells and establishing a human cell atlas.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: There are more microRNAs to be discovered in rare cell types and in developmental stages that have not been assessed.

  • Classified as probably not a substantial claim.

    Interpretationreasoning

    The study is only moving toward a complete cellular microRNAome.

    “Finally, we are only moving toward a complete cellular microRNAome, as many cell types (hepatocytes, neutrophils, pneumocytes) were not available for this study.”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Study design and conductreasoning

    In the Ago CLIP-seq data, distinguishing linker sequences was not necessarily feasible, and 54% of reads were adjusted by CutAdapt to 21- or 22-bp lengths, likely resulting in overidentification of Ago-bound microRNAs.

    “Of note, distinguishing linker sequences in this data set was not necessarily feasible, and 54% of reads were adjusted by CutAdapt to only 21- or 22-bp lengths using the command described above, likely resulting in an overidentification of Ago-bound microRNAs.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Interpretationreasoning

    The reported correlations between superenhancers and microRNAs are not proof of superenhancer activity on the microRNA.

    “We caution that the activity of these superenhancers on adjacent genes and microRNAs is generally unknown, and these reported correlations are not proof of activity of the superenhancer on the microRNA.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing). May combine two claims.

    Study design and conductnone given

    The 159 samples that had less than 1 million microRNA reads were excluded and are not represented in the data.

    “The 159 samples that had less than 1 million microRNA reads were excluded and are not represented in the data.”
    Quote in the paperNot a substantial claim

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Premisecitation

    Expression discretization using DEXUS has been shown to greatly reduce batch effects when combining data across studies and technologies.

    “This type of expression discretization has been shown to greatly reduce batch effects when combining data across studies and technologies (McCall et al. 2011b, 2014).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites McCall et al. 2011b, 2014

      Couldn’t be matched to a paper in Crossref, so it can’t be checked.

  • Probably not a substantial claim: may not pass S3 (load-bearing).

    Study design and conductown data

    Five tissue samples with extreme technical skewing of microRNA reads (>60% of all reads were microRNA let-7b-5p) were removed.

    “We also removed five tissue samples (SRR1635903-8) with extreme technical skewing of microRNA reads (>60% of all reads were microRNA let-7b-5p).”
    Quote in the paperNot a substantial claim
47 of 47

  • The paper relies on this without support.

    MicroRNAs generally act cell-autonomously.

    “MicroRNAs generally act cell-autonomously, and thus their localization to specific cell types is needed to guide our understanding of microRNA activity.”

  • The paper relies on this without support.

    Current tissue-level microRNA data have caused considerable confusion.

    “Current tissue-level data have caused considerable confusion, and comprehensive cell-level data do not yet exist.”

  • The paper relies on this without support.

    Small RNA-seq has become a robust method to fully characterize known microRNAs, capture complete isomiR families, and identify novel microRNAs.

    “Small RNA-seq has become a robust method to fully characterize known microRNAs, capture complete isomiR families, and identify novel microRNAs.”

  • The paper relies on this without support.

    The microRNA community has been forthright in depositing RNA-seq data into central public repositories.

    “The microRNA community has been forthright in depositing RNA-seq data into central public repositories.”

  • The paper relies on this without support.

    Immortalized and cancer-derived cell lines are frequent surrogates for primary cells in understanding biologic pathways.

    “Immortalized and cancer-derived cell lines are frequent surrogates for primary cells in understanding biologic pathways.”

  • The paper relies on this without support.

    The extent of differences in microRNA expression between immortalized/cancer cell lines and primary cells is unknown.

    “However, the extent of differences in microRNA expression between these cell lines and primary cells is unknown.”

  • The paper relies on this without support.

    The cross-contamination of H1264 with H157 is of no consequence for the way data from this cell line were used.

    “However, in the context of the way data from this cell line were used, that distinction is of no consequence here.”

  • The paper relies on this without support.

    There are more microRNAs to be discovered in rare cell types and in developmental stages that have not been assessed.

    “It is likely that there are more microRNAs to be discovered in rare cell types and in developmental stages that have not been assessed.”
8 of 8

Claims other papers make about this one

Each is checked against this paper’s own text: does it say what the citing paper says it does?

  • Many miRNAs, such as the let-7 family, miR-10a, miR-22, and others are also both commonly identified as disease biomarkers and common to many different cell types.

    “Unfortunately, many miRNAs, such as the let-7 family, miR-10a, miR-22, and others are also both commonly identified as disease biomarkers and common to many different cell types [79].”

    From miR-21: a non‐specific biomarker of all maladies (2021) · cites this paper as [79]

    The Results section (‘Diverse microRNAs expression patterns’; page unavailable) reports that several let-7 family members were found in all sampled tissues, and the Abstract discusses why cellular localization matters for candidate miRNA biomarkers. But the paper does not establish that let-7, miR-10a, and miR-22 are all commonly identified disease biomarkers and common to many cell types; it cautions that presence across tissues can reflect a cell-specific miRNA. (LLM, conf=0.85)

    Experimental — accuracy not yet measured

  • Each of these three miRNAs (let-7 family, miR-10a, miR-22) are also implicated in a range of neoplastic and non-neoplastic disorders.

    “Each of these three miRNAs are also implicated in a range of neoplastic and non-neoplastic disorders.”

    From miR-21: a non‐specific biomarker of all maladies (2021) · cites this paper as [79]

    The Results section on diverse microRNA expression patterns reports that several let-7 family members occur across tissues, and the isomiR section discusses miR-10a-5p sequence variation. Neither passage links those miRNAs—or miR-22—to a range of neoplastic and non-neoplastic disorders. Searches for the three miRNAs and disease associations found no passage establishing the claimed links. (LLM, conf=0.74)

    Experimental — accuracy not yet measured

  • Muscle-specific mir-133a-2 is expressed in cardiomyocytes, while mir-148a is expressed in hepatocytes and duct cells and mir-194-2 comes from enterocytes in the gut.

    “we could identify that muscle-specific mir-133a-2 is expressed in cardiomyocytes, while mir-148a is expressed in hepatocytes and duct cells (Fig. 6 C and D) and mir-194–2 comes from the enterocytes in the gut (24, 25).”

    From A mouse tissue atlas of small noncoding RNA (2020) · cites this paper as [24]

  • Well-described brain-specific miRNAs include mir-9, mir-124, mir-219, mir-338, and miRNAs missing from existing catalogs include mir-666, mir-878, mir-433.

    “We identified both well-described brain-specific miRNAs, such as mir-9, mir-124, mir-219, mir-338 (16, 24, 33), as well as those which are missing from existing catalogs, such as mir-666, mir-878, mir-433, etc.”

    From A mouse tissue atlas of small noncoding RNA (2020) · cites this paper as [24]

  • For the majority of identified brain ncRNA, the cell type of origin could not be called based on data generated within previous studies offering cell type resolution.

    “for the majority of these identified RNA we could not call the cell type of origin based in the data generated within previous studies offering cell type resolution (24, 25).”

    From A mouse tissue atlas of small noncoding RNA (2020) · cites this paper as [24]

Self-check

Percentages, sums and reported differences re-computed from the paper’s own numbers.

Each contradiction between two passages listed here was confirmed by an independent check.

  • Nonendothelial cell-type count differs for JHU_ID_23828

    Minor

    The figure caption gives 29 nonendothelial cell types for the comparison, while the Results give 30.

  • Percentage of candidates found in more than 50 samples is incorrect

    Minor

    Using the stated total of 25,218 putative microRNAs, 394 is 1.56%, not 0.7%.

    Reported 0.7 · recomputed 1.562376

  • Percentage of candidates found in individual samples is incorrect

    Minor

    Of the 25,218 putative microRNAs, 18,480 is 73.28%, not 65%.

    Reported 65 · recomputed 73.28099

  • Ago input count changes from 105 to 108

    Minor

    The Methods first identify 105 Ago-precipitated RNA reads, then say 108 Ago reads were processed in miRDeep2.

  • Primary-cell data set count changes from 161 to 162

    Minor

    The Results report 161 primary-cell RNA-seq data sets, whereas the Methods say 162 primary-cell runs were collapsed into the same 46 cell types.

  • Concluding cell-type count conflicts with the reported primary-cell coverage

    Minor

    The conclusion attributes the expression patterns to 42 cell types, but the study reports 46 primary cell types; 42 is elsewhere the count of cancer or immortalized cell lines.

6 of 6

Cites

Flagged first, then by citations

Research works printed but not linked: 1

  1. 2008, Visualizing data using t-SNE, Mach Learn Res, 9, 2579

Cited by

Sorted by citations

Related

Authors
McCall, M. N., Kim, M.-S., Adil, M., Patil, A. H., Lu, Y., Mitchell, C. J., Leal-Rojas, P., Xu, J., Kumar, M., Dawson, V. L., Dawson, T. M., Baras, A. S., Rosenberg, A. Z., Arking, D. E., Burns, K. H., Pandey, A., & Halushka, M. K.
Affiliations
  • University of Rochester Medicine (US)
  • Johns Hopkins University (US)
  • B.S. Abdur Rahman Crescent Institute of Science & Technology (IN)
  • Ginkgo Bioworks, Inc. (United States) (US)
ORCID
Venue
Genome Research
Publisher
Cold Spring Harbor Laboratory
Published
2017-09-06
Volume
27
Issue
10
Pages
1769-1781
ISSN
  • 1088-9051 (ISSN-L)
  • 1549-5469
DOI
10.1101/gr.222067.117
PMC
PMC5630040
OpenAlex
W2951861059
Semantic Scholar
ac6a9702a5d5503965da7115ba3331e833f71007
Cited by
151
References
60
Citations by year
2026: 9 · 2025: 13 · 2024: 8 · 2023: 17 · 2022: 17 · 2021: 19 · 2020: 19 · 2019: 20 · 2018: 26 · 2017: 3
Type
posted content
Language
en
Open access
Locations
  • Genome Research (publishedVersion)
  • PubMed Central (publishedVersion, cc-by-nc)
  • PubMed (publishedVersion)
  • HighWire Press Open Archive (submittedVersion)
  • Omeka S (Universidad de La Frontera) (submittedVersion, cc-by-nc-nd)
Funders
  • American Heart Association (13GRNT16420015)
  • National Institutes of Health
  • Cancer Center Support (P30CA016087)
  • Laura and Isaac Perlmutter Cancer Center
  • NCI's Clinical Proteomic Tumor Analysis Consortium
  • NIH (U24CA160036, U24CA210985)
  • National Fund for Scientific and Technological Development, FONDECYT (1151008, MSCRFII-0429, MSCRFII-0125, 2013-MSCRF-0054, 2014-MSCRF-0665)
  • NIH/NINDS (NS67525, NS37388)
Topics
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Cancer Research › MicroRNA in disease regulation
  • Life Sciences › Agricultural and Biological Sciences › Plant Science › Plant Molecular Biology Research
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Cancer Research › Cancer-related molecular mechanisms research
Keywords
microRNA, human cell types, cancer cell lines, super-enhancers, Drosha, DICER1, small RNA sequencing, colon tissue, adipose tissue, gene expression regulation
MeSH terms
Adult, Cell Line, Transformed, Humans, Male, Organ Specificity, RNA Processing, Post-Transcriptional, MicroRNAs, Cell Line, Tumor

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

Claims read by deepseek-v4.1-flash under definition version 3 (extract-v3) on 2026-11-06.

Classified by jev-1.13-20260917 (gold-guideline-v0.2) on 2026-11-06.

Cited sources are checked by gpt-6-sol.

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review