CoChat
Sign Up
Cited by 777Neuroinflammation and Neurodegeneration MechanismsMultiple Sclerosis Research StudiesSingle-cell and spatial transcriptomics

Transcriptomic analysis of purified human cortical microglia reveals age-associated changes

Open access
Galatro, Thais F, Holtman, Inge R, Lerario, Antonio M, Vainchtein, Ilia D, Brouwer, Nieske, Sola, Paula R, Veras, Mariana M, Pereira, Tulio F +11 more·2017·Nature Neuroscience
DOI 10.1038/nn.4597
Open this paper in CoChat — free account

Abstract

Microglia are essential for CNS homeostasis and innate neuroimmune function, and play important roles in neurodegeneration and brain aging. Here we present gene expression profiles of purified microglia isolated at autopsy from the parietal cortex of 39 human subjects with intact cognition. Overall, genes expressed by human microglia were similar to those in mouse, including established microglial genes CX3CR1, P2RY12 and ITGAM (CD11B). However, a number of immune genes, not identified as part of the mouse microglial signature, were abundantly expressed in human microglia, including TLR, Fcγ and SIGLEC receptors, as well as TAL1 and IFI16, regulators of proliferation and cell cycle. Age-associated changes in human microglia were enriched for genes involved in cell adhesion, axonal guidance, cell surface receptor expression and actin (dis)assembly. Limited overlap was observed in microglial genes regulated during aging between mice and humans, indicating that human and mouse microglia age differently.

AI snapshot

Written from the paper’s full text.

TL;DR

This study generated RNA-seq profiles of FACS-purified microglia from postmortem parietal cortex of 39 cognitively intact human donors and defined a 1,297-gene human microglial core signature. Human microglia broadly resembled mouse microglia but showed human-enriched immune genes and very limited overlap with mouse aging signatures, suggesting that human and mouse microglia age differently.

Key findings

  • A human microglial core signature of 1,297 genes was defined by log fold change > 3 and adjusted P < 0.001 versus cortex; microglia expressed CX3CR1, P2RY12, ITGAM, and TYROBP at high levels, with extremely low/absent CCR2, and were enriched for immune genes including TLR, Fcγ, and SIGLEC receptors plus TAL1 and IFI16.
  • Human microglia overlapped extensively with human CD45+ CNS cells (median Δ percentile 10.1) and with three mouse microglial datasets (median Δ percentiles 14.3, 20.4, and 9.3), but human-enriched genes included CD58, ERAP2, GNLY, and S100A12.
  • With age (donors 34–102 years), 212 genes increased and 360 genes decreased in expression (P < 0.05); the top 100 genes most affected (P < 0.01) separated younger and older samples. Reduced genes included actin cytoskeleton genes TLN1, PFN1, EVL, ARPC1A, ARPC1B, CORO1A, CAP1, CTNNA2, and VASP; sensome receptors P2RY12, IL6R, and TLR10; and adhesion/guidance genes ICAM3, ROBO2, SEMA3C, and SEMA7A. Increased genes included DOCK1, DOCK5, CXCR4, CD163, IGF2R, VEGFA, and RUNX3.
  • Human–mouse aging overlap was limited: 14 increased genes overlapped, including CXCR4, VEGFA, TNFAIP2, and GP2, and 9 decreased genes overlapped, including ETS1, SEMA7A, MRC2, PSTPIP1, and EMP2; about half of the shared aging genes showed opposite directionality between species.

Methods

  • Microglia were acutely isolated by FACS from right parietal cortex autopsy tissue; 81 rapid isolations were performed (33 in the Netherlands, 48 in Brazil), and 39 high-quality RNA samples from donors with normal cognition and no apparent neuropathological abnormalities were RNA-sequenced, alongside 16 superior parietal cortex samples and 10 epilepsy surgery biopsies. Protein samples came from 7 donors.
  • Paired-end Illumina HiSeq 2500 RNA-seq data were analyzed with limma, with batch/unwanted variation removed using sva; PCA assessed sample segregation, and differential expression used adjusted P < 0.001 and log(FC) > 3 for the microglial core signature. GO enrichment used GOstats, transcriptional regulators used CoRegNet, and protein interactions used STRING.
  • Protein validation used western blotting, human hematopoietic antibody arrays on pooled FACS-sorted microglia, and immunohistochemistry/double immunofluorescence for IBA1, HLA-DR, CD32, CD36, LAG3, and TMEM119.

Limitations

  • Stated: all donors suffered from pathologies ultimately leading to death, which may have affected microglial gene expression; human autopsy samples were compared with healthy mouse microglia; monocyte/macrophage comparisons may be partly confounded by batch effects; the sample was unbalanced by gender (7 females, 32 males), so more female donors would be needed to identify gender-specific traits; neuronal mRNAs in aged microglia might reflect phagocytosis during postmortem delay or age-related neuronal loss rather than intrinsic microglial expression.
  • Evident: cross-species comparisons relied on different technologies and platforms (3′-end RNA sequencing, microarrays, RNA-seq) and small mouse sample sizes (n = 6, n = 4, n = 2 in the cited datasets), and the aging analyses are correlational, so they cannot establish causal effects of age on microglial function.

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

Integrity signals

Registry

Crossref · 10.1038/nn.4597

Cites

62 of 62 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Claims the paper makes

0 of 37 cited sources checked · 0 supported · 0 flagged

  • Findingown data
    § DISCUSSION · p. 10

    Using CoRegNet, the study identified regulators of the human core microglial genes.

    “Using CoRegNet, we identified regulators of the human core microglial genes.”
    Quote in the paperNot a substantial claim
  • Findingown data
    § DISCUSSION · p. 10

    These transcription factors (TAL1, RUNX1, LYL1) are abundantly expressed in the human microglial core.

    “and these transcription factors are abundantly expressed in the human microglial core.”
    Quote in the paperNot a substantial claim
  • Findingown data
    § Identification of the core human microglial gene signature · p. 3

    When microglial expression data were analyzed using PMD as a variable, no effect was observed.

    “When we analyzed the microglial expression data using PMD as a variable, we observed no effect.”
    Quote in the paperNot a substantial claim
  • Findingown data
    § DISCUSSION · p. 10

    The study observed very little overlap between the human and mouse aging signatures.

    “Strikingly, we observed very little overlap between the human and mouse aging signatures.”
    Quote in the paperSubstantial claim
  • Findingown data
    § DISCUSSION · p. 10

    The study observed age-associated changes in genes involved in axonal guidance, cell adhesion and the sensome.

    “In addition, we observed age-associated changes in genes involved in axonal guidance, cell adhesion and the sensome.”
    Quote in the paperSubstantial claim
  • May combine two claims. 28
  • The quote only partly states this claim. 2
  • Classified as probably not a substantial claim. 44

  • May combine two claims.

    Findingown data
    § Age-associated changes in microglial gene expression · p. 9

    Genes with higher expression during aging included integrin modulators DOCK1 and DOCK5, receptors CXCR4, CD163 and IGF2R, growth factor VEGFA and transcription factor RUNX3.

    “Genes with higher expression during aging included integrin modulators DOCK1 and DOCK5 , receptors CXCR4 , CD163 and IGF2R , growth factor VEGFA and transcription factor RUNX3 .”
    Quote in the paperSubstantial claim
    • Rests on an assumption: Age-associated gene expression changes in the donor microglia are attributable to aging rather than to cause of death, comorbidity or other donor characteristics.

  • May combine two claims.

    Findingown data
    § Age-associated changes in microglial gene expression · p. 9

    Genes with reduced expression during aging included actin cytoskeleton-associated genes TLN1, PFN1, EVL, ARPC1A, ARPC1B, CORO1A, CAP1, CTNNA2 and VASP, sensome cell surface receptors P2RY12, IL6R and TLR10, and cell adhesion molecules and cell surface receptors ICAM3, ROBO2, SEMA3C and SEMA7A.

    “Noteworthy genes with reduced expression during aging included many actin cytoskeleton-associated genes such as TLN1 , PFN1 , EVL , ARPC1A , ARPC1B , CORO1A , CAP1 , CTNNA2 and VASP , sensome cell surface receptors P2RY12 , IL6R and TLR10 , and cell adhesion molecules and cell surface receptors ICAM3 , ROBO2 , SEMA3C and SEMA7A .”
    Quote in the paperSubstantial claim
    • Rests on an assumption: Age-associated gene expression changes in the donor microglia are attributable to aging rather than to cause of death, comorbidity or other donor characteristics.

  • May combine two claims.

    Findingown data
    § Overlap between human and mouse microglial gene expression profiles · p. 4

    There was extensive overlap in gene expression between the human microglial core and the published human CD45+ microglial cell population; the median Δ percentile was 10.1.

    “Overall, we observed an extensive overlap in gene expression between these data sets: the median ∆ percentile was 10.1, indicating most genes were similarly expressed in both data sets ( Fig. 2b ).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [15] Zhang, Y. et al. Purification and characterization of progenitor and mature human astrocytes reveals transcriptional and functional differences with mouse. Neuron 89, 37–53 (2016). source

      Queued — the background check reaches it, usually within a day.

    • Rests on an assumption: The published human CD45+ cell population from Zhang et al. is comparable to microglia for the overlap analysis.

  • The quote only partly states this claim.

    Findingown data
    § DISCUSSION · p. 10

    Of the small number of genes shared between human and mouse during aging, about half displayed opposite directionalities in expression during aging.

    “Of the small number of genes that were shared between human and microglia during aging, about half displayed opposite directionalities in expression during aging-for example, upregulated in mouse and downregulated in human or vice versa.”
    Quote in the paperSubstantial claim

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

    Findingnone given
    § DISCUSSION · p. 10

    Genes involved in cell cycle are also highly expressed in human microglia.

    “while genes involved in cell cycle are also highly expressed.”
    Quote in the paperNot a substantial claim

  • The quote only partly states this claim.

    Findingown data
    § Overlap between human and mouse microglial gene expression profiles · p. 5

    Several human microglial genes were much less abundantly expressed in mouse microglia, including CD58, ERAP2, GNLY and S100A12.

    “Notably, several human microglial genes were much less abundantly expressed in mouse microglia. Examples of human microgliaenriched genes are CD58, an adhesion molecule with a central role in the clustering of mature dendritic cells and T lymphocyte activation 25 .”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data
    § Overlap between human and mouse microglial gene expression profiles · p. 5

    PCA of human microglial and published monocyte and macrophage data indicated that these three data sets segregated and that microglia were the most distinct.

    “PCA of our human microglial and published monocyte and macrophage data 30 indicated that these three data sets segregated and that microglia were the most distinct ( Supplementary Figure 1a ).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [30] Saeed, S. et al. Epigenetic programming of monocyte-to-macrophage differentiation and trained innate immunity. Science 345, 1251086 (2014). source

      Queued — the background check reaches it, usually within a day.

  • May combine two claims.

    Findingown data
    § Overlap between human and mouse microglial gene expression profiles · p. 5

    An extensive overlap was observed between human microglia and three independent mouse microglial gene expression data sets; median Δ percentiles were 14.3, 20.4 and 9.3.

    “An extensive overlap was observed between human and these three independent mouse microglial gene expression data sets ( Fig. 2c ). The median ∆ percentiles were 14.3, 20.4 and 9.3 for our own expression data set, that of Grabert et al. 23 and that of Matcovich-Natan et al. 24 , respectively.”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites [23] Grabert, K. et al. Microglial brain region-dependent diversity and selective regional sensitivities to aging. Nat. Neurosci. 19, 504–516 (2016). source

      Queued — the background check reaches it, usually within a day.

    • cites [24] Matcovitch-Natan, O. et al. Microglia development follows a stepwise program to regulate brain homeostasis. Science 353, aad8670 (2016). source
    • Rests on an assumption: Cross-species comparisons using different RNA quantification technologies (3' QuantSeq, microarrays and RNA-seq) are valid for assessing overlap.

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

    Findingnone given
    § DISCUSSION · p. 10

    ID4 is downregulated in microglia.

    “as exemplified by the downregulation of ID4 in microglia.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data
    § Identification of the core human microglial gene signature · p. 3

    Microglial and cortical samples segregated cleanly in PCA and were highly similar within each tissue group.

    “Microglial and cortical samples segregated cleanly and were highly similar within each tissue group, indicating that microglial expression profiles are very different from unsorted parietal cortex expression data ( Fig. 1b ).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Findingcitation
    § DISCUSSION · p. 10

    TLN1, PFN1, and VASP are less abundant in microglia isolated from older donors.

    “For instance, talin-1 (TLN1), a protein involved in actin remodeling and cell motility; profilin (PFN1), a protein directly involved in actin binding and polymerization; and VASP, a protein involved in actin filaments elongation, are less abundant in microglia isolated from older donors 47 .”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [47] Vincent, C., Siddiqui, T.A. & Schlichter, L.C. Podosomes in migrating microglia: components and matrix degradation. J. Neuroinflammation 9, 190 (2012). source

      Queued — the background check reaches it, usually within a day.

  • Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based).

    Findingnone given
    § DISCUSSION · p. 10

    Human microglia seem to possess a highly activated pathway for proliferation (the CXCR-CD74 axis).

    “Notably, although human microglia seem to possess a highly activated pathway for proliferation (the aforementioned CXCR-CD74 axis),”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data
    § Microglial protein expression · p. 9

    Cells with highly ramified microglial morphology were detected with antibodies against IBA1, HLA-DR, CD32, CD36, LAG3 and TMEM119, and extensive colocalization with IBA1 and HLA-DR was observed.

    “Cells with highly ramified microglial morphology were detected with the indicated antibodies. Using double immunofluorescence staining, we observed extensive colocalization with established microglial markers IBA1 and HLA-DR.”
    Quote in the paperNot a substantial claim

  • May combine two claims.

    Findingown data
    § DISCUSSION · p. 9

    Considerable overlap between human and mouse microglial gene expression was observed, but noteworthy dissimilarities were also present.

    “Considerable overlap between human and mouse microglial gene expression was observed, but noteworthy dissimilarities were also present.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: Cross-species comparisons using different RNA quantification technologies (3' QuantSeq, microarrays and RNA-seq) are valid for assessing overlap.

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

    Premisecitation
    § DISCUSSION · p. 10

    NLRC5 is constitutively expressed in various tissues, most highly in cells from the hematopoietic lineage (CD4+ T cells, CD8+ T cells, CD19+ B cells, natural killer cells and natural killer T cells).

    “While CIITA expression is restricted to antigen-presenting cells, NLRC5 is constitutively expressed in various tissues, most highly in cells from the hematopoietic lineage (CD4 + T cells, CD8 + T cells, CD19 + B cells, natural killer cells and natural killer T cells).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [39] Meissner, T.B. et al. NLR family member NLRC5 is a transcriptional regulator of MHC class I genes. Proc. Natl. Acad. Sci. USA 107, 13794–13799 (2010). source

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

    Study design and conductnone given
    § ONLINE METHODS · p. 12

    Written informed consent was obtained and ethics committees approved the procedure.

    “Written informed consent for research purposes was obtained and ethics committees from the School of Medicine of the University of São Paulo, Brazil, the Netherlands Brain Bank and the University Medical Center Groningen, the Netherlands approved the procedure.”
    Quote in the paperNot a substantial claim

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

    Premisecitation
    § DISCUSSION · p. 10

    In adults, TAL1 is expressed in hematopoietic cells but also in specific regions of the midbrain, hindbrain and spinal cord.

    “In adults, TAL1 is expressed in hematopoietic cells but also in specific regions of the midbrain, hindbrain and spinal cord 41 .”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [41] van Eekelen, J.A. et al. Expression pattern of the stem cell leukaemia gene in the CNS of the embryonic and adult mouse. Neuroscience 122, 421–436 (2003). source

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

    Study design and conductcitation
    § ONLINE METHODS · p. 12

    FAST-QC quality checks were done to determine high quality base pairs.

    “FAST-QC (http://www.bioinformatics.babraham.ac.uk/projects/fastqc/) 52 quality checks were done to determine high quality base pairs,”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [52] Babraham Bioinformatics. FastQC: a quality control tool for high throughput sequence data http://www.bioinformatics.babraham.ac.uk/projects/fastqc/ (accessed 5 February 2017).

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

  • May combine two claims.

    Interpretationown data
    § Identification of the core human microglial gene signature · p. 4

    Human microglia are immunocompetent and phagocytic cells of the CNS that express a wide range of immune receptors and ligands.

    “These data confirm that human microglia are immunocompetent and phagocytic cells of the CNS that express a wide range of immune receptors and ligands, equipped to respond to a wide variety of pathogen- and damage-associated molecules.”
    Quote in the paperNot a substantial claim
    • Rests on an assumption: The 39 selected samples are representative of normal human cortical microglia.

  • Classified as probably not a substantial claim.

    Study design and conductnone given
    § ONLINE METHODS · p. 12

    Cells were sorted on the basis of CD11b high CD45 int expression and negative staining for DAPI or LIVE/DEAD.

    “Cells were sorted on the basis of CD11b high CD45 int expression and negative staining for DAPI or the LIVE/DEAD Fixable Red Dead Cell stain (L23102, ThermoFisher Scientific) and collected in RNAlater (Qiagen, Germany).”
    Quote in the paperSubstantial claim

  • May combine two claims.

    Interpretationown data
    § Age-associated changes in microglial gene expression · p. 9

    The overlap in genes differentially expressed during aging between human and mouse is very limited, suggesting that microglia of physiologically aged mice do not recapitulate the effect of aging on human microglia.

    “The overlap in genes differentially expressed during aging between human and mouse is very limited, suggesting that the microglia of physiologically aged mice do not recapitulate the effect of aging on human microglia.”
    Quote in the paperSubstantial claim
    • Rests on an assumption: Age-associated gene expression changes in the donor microglia are attributable to aging rather than to cause of death, comorbidity or other donor characteristics.

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

    Premisecitation
    § DISCUSSION · p. 10

    A few studies have identified a gender-specific role for microglia in the rat brain.

    “Little is known about microglial gender differences, but a few studies have identified a gender-specific role for microglia in the rat brain 49 .”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [49] Chen, Y., Won, S.J., Xu, Y. & Swanson, R.A. Targeting microglial activation in stroke therapy: pharmacological tools and gender effects. Curr. Med. Chem. 21, 2146–2155 (2014). source

  • Classified as probably not a substantial claim.

    Study design and conductnone given
    § ONLINE METHODS · p. 12

    The leptomeninges were removed to avoid potential contamination by meningeal macrophages.

    “The leptomeninges were removed to avoid potential contamination by meningeal macrophages.”
    Quote in the paperSubstantial claim

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

    Premisecitation
    § DISCUSSION · p. 10

    Recent reports on the induction of human pluripotent stem cell-derived microglia-like cells demonstrated high levels of IRF8 in such cells.

    “recent reports on the induction of human pluripotent stem cell-derived microglia-like cells demonstrated high levels of this transcription factor in such cells 40 .”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [40] Muffat, J. et al. Efficient derivation of microglia-like cells from human pluripotent stem cells. Nat. Med. 22, 1358–1367 (2016). source

      Not checked in the free run.

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

    Premisecitation
    § DISCUSSION · p. 10

    SIGLECs regulate the effects of activation, phagocytosis and inflammasome formation via their immunoreceptor tyrosine-based inhibition motif (ITIM).

    “SIGLECs regulate the effects of activation, phagocytosis and inflammasome formation via their immunoreceptor tyrosine-based inhibition motif (ITIM) 37 .”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [37] Macauley, M.S., Crocker, P.R. & Paulson, J.C. Siglec-mediated regulation of immune cell function in disease. Nat. Rev. Immunol. 14, 653–666 (2014). source

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

    Study design and conductown data
    § ONLINE METHODS · p. 12

    A quality check of aligned data was done with RNA-SEQC and showed high quality alignment and data.

    “A quality check of aligned data was done with RNA-SEQC 55 and showed high quality alignment and data.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [55] DeLuca, D.S. et al. RNA-SeQC: RNA-seq metrics for quality control and process optimization. Bioinformatics 28, 1530–1532 (2012). source

      Not checked in the free run.

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

    Premisecitation
    § DISCUSSION · p. 10

    These receptors and their downstream signaling are implicated in chemotaxis and are part of the microglial sensome.

    “These receptors and their downstream signaling are implicated in chemotaxis and are part of the microglial sensome 4 .”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [4] Hickman, S.E. et al. The microglial sensome revealed by direct RNA sequencing. Nat. Neurosci. 16, 1896–1905 (2013). source

      Not checked in the free run.

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

    Premisecitation
    § DISCUSSION · p. 10

    RUNX1 activates the transcription of SPI-1, a bona fide microglial transcription factor.

    “RUNX1 activates the transcription of SPI-1 (ref. 43), a bona fide microglial transcription factor.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [43] Huang, G. et al. PU.1 is a major downstream target of AML1 (RUNX1) in adult mouse hematopoiesis. Nat. Genet. 40, 51–60 (2008). source

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

    Premisecitation
    § DISCUSSION · p. 10

    P2RY12 is an established microglial marker.

    “P2RY12, an established microglial marker 3,4 ,”
    Quote in the paperNot a substantial claim0 of 2 sources checked
    • cites [3] Butovsky, O. et al. Identification of a unique TGF-β-dependent molecular and functional signature in microglia. Nat. Neurosci. 17, 131–143 (2014). source

      Not checked in the free run.

    • cites [4] Hickman, S.E. et al. The microglial sensome revealed by direct RNA sequencing. Nat. Neurosci. 16, 1896–1905 (2013). source

      Not checked in the free run.

  • Classified as probably not a substantial claim.

    Premisecitation
    § DISCUSSION · p. 10

    CIITA expression is restricted to antigen-presenting cells.

    “While CIITA expression is restricted to antigen-presenting cells, NLRC5 is constitutively expressed in various tissues, most highly in cells from the hematopoietic lineage (CD4 + T cells, CD8 + T cells, CD19 + B cells, natural killer cells and natural killer T cells).”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [39] Meissner, T.B. et al. NLR family member NLRC5 is a transcriptional regulator of MHC class I genes. Proc. Natl. Acad. Sci. USA 107, 13794–13799 (2010). source

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

    Premisecitation
    § DISCUSSION · p. 9

    Host microbiota control the maturation and function of microglia.

    “This is supported by observations in mice, where host microbiota control the maturation and function of microglia 33 .”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [33] Erny, D. et al. Host microbiota constantly control maturation and function of microglia in the CNS. Nat. Neurosci. 18, 965–977 (2015). source

      Not checked in the free run.

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

    Premisecitation
    § DISCUSSION · p. 10

    NLRC5 expression levels are intermediate in monocytic CD14+ cells and CD11B+ splenic myeloid cells.

    “NLRC5 expression levels are intermediate in monocytic CD14 + cells and CD11B + splenic myeloid cells.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [39] Meissner, T.B. et al. NLR family member NLRC5 is a transcriptional regulator of MHC class I genes. Proc. Natl. Acad. Sci. USA 107, 13794–13799 (2010). source

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

    Study design and conductnone given
    § ONLINE METHODS · p. 12

    RNA quality was checked with Experion RNA HighSens Analysis Kit and RNA Screen Tape.

    “RNA quality was checked with the Experion RNA HighSens Analysis Kit (Bio-Rad, USA) and RNA Screen Tape (Agilent Technologies, USA).”
    Quote in the paperNot a substantial claim

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

    Premisecitation
    § DISCUSSION · p. 10

    IFI16 (p204) is associated with viral infection and the response to cytosolic double-stranded DNA.

    “IFI16 (p204) is associated with viral infection and the response to cytosolic double-stranded DNA 44 .”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [44] Gürtler, C. & Bowie, A.G. Innate immune detection of microbial nucleic acids. Trends Microbiol. 21, 413–420 (2013). source

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

    Premisecitation
    § DISCUSSION · p. 10

    CD33 (SIGLEC3) is associated with accumulation of amyloidβ plaques in mouse Alzheimer's disease models due to diminished phagocytic capacity of microglial cells upregulating this gene.

    “CD33 (SIGLEC3), a known microglial marker, is associated with the accumulation of amyloidβ plaques in mouse Alzheimer's disease models due to the diminished phagocytic capacity of microglial cells upregulating this gene 38 .”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [38] Griciuc, A. et al. Alzheimer's disease risk gene CD33 inhibits microglial uptake of amyloid beta. Neuron 78, 631–643 (2013). source

      Not checked in the free run.

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

    Premisecitation
    § DISCUSSION · p. 10

    IRF8 is necessary for microglial genesis, along with SPI-1.

    “IRF8 is necessary for microglial genesis, along with SPI-1 (ref. 2),”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [2] Kierdorf, K. et al. Microglia emerge from erythromyeloid precursors via Pu.1- and Irf8-dependent pathways. Nat. Neurosci. 16, 273–280 (2013). source

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

    Premisecitation
    § Identification of the core human microglial gene signature · p. 4

    In monocytes, heterodimers between CXCR2 and CD74 or between CXCR4 and CD74 induce PI3K-AKT signaling and p53 inhibition, leading to increased survival, proliferation and apoptosis evasion.

    “In monocytes, heterodimers between CXCR2 and CD74 or between CXCR4 and CD74 induce PI3K-AKT signaling and p53 inhibition, leading to increased survival, proliferation and apoptosis evasion 22 .”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [22] Mitchell, R.A. et al. Macrophage migration inhibitory factor (MIF) sustains macrophage proinflammatory function by inhibiting p53: regulatory role in the innate immune response. Proc. Natl. Acad. Sci. USA 99, 345–350 (2002). source

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

    Interpretationreasoning
    § DISCUSSION · p. 10

    Human microglia are able to interact with CD4 and CD8 T-helper cell subsets.

    “These data corroborate the notion that human microglia are potent antigen-presenting cells and are able to interact with CD4 and CD8 T-helper cell subsets.”
    Quote in the paperNot a substantial claim

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

    Premisecitation
    § DISCUSSION · p. 10

    IFI16 regulates cell proliferation and transcription through different interactions with p53 and STAT3.

    “It regulates cell proliferation and transcription through different interactions with p53 and STAT3.”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [44] Gürtler, C. & Bowie, A.G. Innate immune detection of microbial nucleic acids. Trends Microbiol. 21, 413–420 (2013). source

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

    Premisecitation
    § Identification of the core human microglial gene signature · p. 4

    Integrin-SYK signaling negatively regulates TLR signaling and the NFκB pathway by promoting degradation of MYD88.

    “Integrin-SYK signaling is also responsible for negatively regulating Toll-like receptor (TLR) signaling and the NFκ B pathway by promoting the degradation of MYD88 (ref. 21).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [21] Kondo, T., Kawai, T. & Akira, S. Dissecting negative regulation of Toll-like receptor signaling. Trends Immunol. 33, 449–458 (2012). source

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

    Study design and conductnone given
    § ONLINE METHODS · p. 12

    No statistical methods were used to predetermine sample sizes; as many samples as possible were collected in the available time period.

    “No statistical methods were used to predetermine samples sizes; as many samples as possible were collected in the available time period.”
    Quote in the paperNot a substantial claim

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

    Premisecitation
    § Transcription factors associated with human microglial core gene expression · p. 7

    SPI-1 (PU.1) and IRF8 are essential for microglial ontogeny.

    “Two of these transcription factors are essential for microglial ontogeny: SPI-1 (PU.1) and IRF8 (ref. 2).”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [2] Kierdorf, K. et al. Microglia emerge from erythromyeloid precursors via Pu.1- and Irf8-dependent pathways. Nat. Neurosci. 16, 273–280 (2013). source

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

    Interpretationreasoning
    § DISCUSSION · p. 10

    IFI16 may support the relative microglial quiescence that is observed under normal conditions.

    “IFI16 may support the relative microglial quiescence that is observed under normal conditions.”
    Quote in the paperNot a substantial claim
    • Rests on an assumption: The CXCR-CD74 pathway is dependent on extracellular signals.

  • May combine two claims.

    Interpretationown data
    § DISCUSSION · p. 10

    Human microglia express many SIGLECs, highlighting their importance and the specialization of microglia for maintenance of a homeostatic microenvironment.

    “Our finding that human microglia express many SIGLECs highlights their importance and the specialization of microglia for the maintenance of a homeostatic microenvironment.”
    Quote in the paperNot a substantial claim
    • Rests on an assumption: The 39 selected samples are representative of normal human cortical microglia.

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

    Premisecitation
    § DISCUSSION · p. 10

    Chromatin immunoprecipitation and sequencing analysis previously indicated that TAL1 forms a complex with RUNX1 and LYL1.

    “Chromatin immunoprecipitation and sequencing analysis previously indicated that TAL1 forms a complex with RUNX1 and LYL1 (ref. 42),”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [42] Wilson, N.K. et al. Combinatorial transcriptional control in blood stem/progenitor cells: genome-wide analysis of ten major transcriptional regulators. Cell Stem Cell 7, 532–544 (2010). source

      Not checked in the free run.

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

    Premisecitation
    § DISCUSSION · p. 10

    Sex differences in microglial function in the developing brain are associated with neurodevelopmental and psychiatric conditions.

    “Sex differences in microglial function in the developing brain are associated with neurodevelopmental and psychiatric conditions 50 .”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [50] Lenz, K.M. & McCarthy, M.M. A starring role for microglia in brain sex differences. Neuroscientist 21, 306–321 (2015). source

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

    Premisecitation
    § DISCUSSION · p. 10

    NLRC5 and CIITA are the respective master regulators of MHC class I and II gene expression.

    “NLRC5 and CIITA are the respective master regulators of MHC class I and II gene expression 39 .”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [39] Meissner, T.B. et al. NLR family member NLRC5 is a transcriptional regulator of MHC class I genes. Proc. Natl. Acad. Sci. USA 107, 13794–13799 (2010). source

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

    Premisecitation
    § DISCUSSION · p. 10

    IFI16 is involved in overcoming inhibitor of differentiation proteins (IDs) blocking of terminal cell differentiation.

    “IFI16 is also involved in overcoming inhibitor of differentiation proteins (IDs) blocking of terminal cell differentiation 45 ,”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [45] Luan, Y., Lengyel, P. & Liu, C.-J. p204, a p200 family protein, as a multifunctional regulator of cell proliferation and differentiation. Cytokine Growth Factor Rev. 19, 357–369 (2008). source

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

    Premisecitation
    § DISCUSSION · p. 10

    Microglial process motility directed toward site of damage was significantly impaired in mice with Aβ plaques.

    “Notably, microglial process motility directed toward site of damage was significantly impaired in mice with A β plaques 48 .”
    Quote in the paperNot a substantial claim0 of 1 sources checked
    • cites [48] Krabbe, G. et al. Functional impairment of microglia coincides with Beta-amyloid deposition in mice with Alzheimer-like pathology. PLoS One 8, e60921 (2013). source

      Not checked in the free run.

  • Classified as probably not a substantial claim.

    Study design and conductcitation
    § DISCUSSION · p. 9

    To minimize the reported effect of brain region on microglial gene expression, only cortical microglia were included in comparisons.

    “To minimize the reported effect of brain region on microglial gene expression 6,23,24 , only cortical microglia were included in our comparisons.”
    Quote in the paperSubstantial claim0 of 3 sources checked
    • cites [6] Zhang, Y. et al. An RNA-sequencing transcriptome and splicing database of glia, neurons, and vascular cells of the cerebral cortex. J. Neurosci. 34, 11929–11947 (2014). source

      Queued — the background check reaches it, usually within a day.

    • cites [23] Grabert, K. et al. Microglial brain region-dependent diversity and selective regional sensitivities to aging. Nat. Neurosci. 19, 504–516 (2016). source

      Not checked in the free run.

    • cites [24] Matcovitch-Natan, O. et al. Microglia development follows a stepwise program to regulate brain homeostasis. Science 353, aad8670 (2016). source

  • Classified as probably not a substantial claim.

    Study design and conductnone given
    § ONLINE METHODS · p. 12

    The expression values were adjusted according to the surrogate variables identified by sva using removeBatchEffect.

    “The expression values were adjusted according to the surrogate variables identified by sva using the function removeBatchEffect from the limma package.”
    Quote in the paperSubstantial claim

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

    Study design and conductown data
    § ONLINE METHODS · p. 12

    Per sample, on average 17.6 million uniquely mapped reads were generated, ranging between 6 and 61.1 million.

    “Per sample, on average 17.6 million uniquely mapped reads were generated, ranging between 6 and 61.1 million.”
    Quote in the paperNot a substantial claim

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

    Study design and conductnone given
    § ONLINE METHODS · p. 12

    The second comparison was performed as a multivariate linear model that included gender as a binary variable, while age, PMD and surrogate variables were used as quantitative variables.

    “The second comparison was performed as a multivariate linear model that included gender as a binary variable, while age, PMD and the surrogate variables calculated with sva were used as quantitative variables.”
    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
    § ONLINE METHODS · p. 12

    Samples were not randomized; all collected samples were RNA-sequenced or used for protein sample preparation.

    “Samples were not randomized; all collected samples were RNA-sequenced or used for protein sample preparation.”
    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
    § ONLINE METHODS · p. 12

    Data collection and analysis were not performed blind to the conditions of the experiments and no data points were excluded from the analyses.

    “Data collection and analysis were not performed blind to the conditions of the experiments and no data points were excluded from the analyses.”
    Quote in the paperNot a substantial claim

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

    Study design and conductnone given
    § ONLINE METHODS · p. 13

    Vimentin was used as an internal standard for membrane normalization.

    “Vimentin was used as in internal standard for membrane normalization.”
    Quote in the paperNot a substantial claim

  • Classified as probably not a substantial claim.

    Study design and conductcitation
    § ONLINE METHODS · p. 12

    Unwanted or hidden sources of variation, such as the use of different library preps and different batches, were removed using the sva package.

    “Unwanted or hidden sources of variation, such as the use of different library preps (Illumina and Takara) and different batches, were removed using the sva package 60 .”
    Quote in the paperSubstantial claim0 of 1 sources checked
    • cites [60] Leek, J.T., Johnson, W.E., Parker, H.S., Jaffe, A.E. & Storey, J.D. The sva package for removing batch effects and other unwanted variation in high-throughput experiments. Bioinformatics 28, 882–883 (2012). source

      Not checked in the free run.

  • Probably not a substantial claim: may not pass S3 (load-bearing), S4 (based). May combine two claims.

    Study design and conductnone given
    § ONLINE METHODS · p. 12

    The distribution of the RNA-seq data was as expected for sequencing data; no particular testing was performed.

    “The distribution of the RNA-seq data was as expected for sequencing data; no particular testing was performed.”
    Quote in the paperNot a substantial claim
58 of 58

  • The paper relies on this without support.

    The pathologies that led to the donors' deaths did not substantially alter the microglial transcriptome, so the profiles approximate physiological human microglia.

    “Nonetheless, all donors suffered from a range of pathologies ultimately leading to their death, potentially affecting the microglial gene expression profile.”

  • The paper relies on this without support.

    Age-associated gene expression changes in the donor microglia are attributable to aging rather than to cause of death, comorbidity or other donor characteristics.

    “The ages of our donors ranged between 34 and 102 years, allowing us to examine changes in microglial gene expression that occurred during aging.”

  • The paper relies on this without support.

    The 39 selected samples are representative of normal human cortical microglia.

    “Ultimately, 39 samples from donors with a history of normal cognitive function and no apparent neuropathological abnormalities with sufficient levels of high-quality RNA were selected for RNA sequencing.”

  • The paper relies on this without support.

    Cross-species comparisons using different RNA quantification technologies (3' QuantSeq, microarrays and RNA-seq) are valid for assessing overlap.

    “Even though these mouse expression data were generated using very different technologies (3 ′ -end polyadenylated RNA sequencing, microarrays and RNA-seq, respectively), we observed an extensive overlap with human microglial data.”

  • The paper relies on this without support.

    The published human CD45+ cell population from Zhang et al. is comparable to microglia for the overlap analysis.

    “We compared our human microglial expression profile to a published profile of the human CD45 + microglial cell population, isolated using immunopanning from temporal lobe cortex tissue obtained during tumor or epilepsy surgery 15 .”

  • The paper relies on this without support.

    The CXCR-CD74 pathway is dependent on extracellular signals.

    “this pathway is dependent on extracellular signals,”

  • The paper relies on this without support.

    Actin (dis)assembly is essential for cellular motility and migration.

    “As actin (dis)assembly is essential for cellular motility and migration,”

  • The paper relies on this without support.

    Binding of chemoattractant ligands to their respective receptors on myeloid cells induces these cells to move toward injury sites, apoptotic cells or invading microorganisms.

    “Binding of chemoattractant ligands to their respective receptors on myeloid cells induces these cells to move toward injury sites, apoptotic cells or invading microorganisms.”

  • The paper relies on this without support.

    Movement of fine microglial processes to sense the environment and initiate chemotaxis is primarily governed through these receptors.

    “Movement of fine microglial processes to sense the environment and initiate chemotaxis is primarily governed through these receptors,”
9 of 9

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?

  • Other studies have questioned the overall degree of conservation between AD mouse models and human brains.

    “Although some studies have identified selected overlaps in expression changes between AD mouse models and human brains (Bennett et al., 2018; Castillo et al., 2017; Mostafavi et al., 2018; Neuner et al., 2019; Rojo et al., 2017), other studies have questioned the overall degree of conservation (Burns et al., 2015; Galatro et al., 2017; Hargis and Blalock, 2017).”

    From Meta-Analysis of the Alzheimer’s Disease Human Brain Transcriptome and Functional Dissection in Mouse Models (2020) · cites this paper as Galatro et al., 2017

    The Discussion (p. 10) reports very little overlap between human and mouse microglial aging signatures, supporting concern about cross-species conservation. But the paper compares microglia during aging, not AD mouse models with human AD brains; it does not establish the claim at that disease-specific level. (LLM, conf=0.86)

    Experimental — accuracy not yet measured

  • Animal microglia are often inaccessible to real-time imaging and cannot fully reflect human biology.

    “Animal microglia are often inaccessible to real-time imaging, and cannot fully reflect human biology.15–17”

    From Tubular human brain organoids to model microglia-mediated neuroinflammation (2021) · cites this paper as [17]

    The Discussion (p. 10) reports critical differences between human and mouse microglia, especially with aging, supporting the point that animal microglia cannot fully reflect human biology. The paper does not say that animal microglia are often inaccessible to real-time imaging; a search for real-time and in vivo imaging found no support for that part. (LLM, conf=0.88)

    Experimental — accuracy not yet measured

  • Recent transcriptome-wide studies of bulk ex vivo human microglia have consistently suggested that microglia change with age and have transcriptomes enriched for disease-related genes.

    “Recent transcriptome-wide studies of bulk ex vivo human microglia have consistently suggested that microglia change with age and have transcriptomes that are enriched for disease-related genes3–5.”

    From Single cell RNA sequencing of human microglia uncovers a subset associated with Alzheimer’s disease (2020) · cites this paper as [3]

    The ‘Age-associated changes in microglial gene expression’ section (p. 9) reports expression changes across human donors aged 34–102, supporting the aging part of the claim. The ‘Identification of the core human microglial gene signature’ section (p. 4) reports enrichment for immune-response and other functional terms, but does not establish enrichment for disease-related genes. (LLM, conf=0.85)

    Experimental — accuracy not yet measured

  • Genes involved in actin assembly normally required for morphogenic changes in microglia are downregulated during aging, suggesting fundamental microglial functions are impaired with age.

    “Recent transcriptomic analysis of human cortical microglia implicates that genes involved in actin assembly normally required to mediate morphogenic changes in microglia are downregulated during aging. As dynamic filamentous actin (F-actin) assembly is essential to microglia morphogenesis, migration, and Aβ uptake/clearance, these results suggest that fundamental microglial functions are impaired”

    From Molecular and cellular mechanisms underlying the pathogenesis of Alzheimer’s disease (2020) · cites this paper as [523]

  • There was surprisingly limited overlap in microglial genes regulated by aging in human versus mouse.

    “There was surprisingly limited overlap, however, in microglial genes regulated by aging in human versus mouse (Grabert et al., 2016; Galatro et al., 2017)”

    From Microglia in Alzheimer’s disease (2018) · cites this paper as Galatro et al., 2017

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.

  • Postmortem-delay range differs

    Minor

    The results state that microglia were isolated after a 6- to 24-hour postmortem delay, while the discussion describes the assessed range as 4 to 24 hours.

  • Percentile-comparison formula does not define the reported difference

    Minor

    The results define the change in percentile by subtracting the two percentiles. The methods instead specify one minus the absolute difference, which is a different quantity.

  • CoRegNet results assigned to different figures

    Minor

    The results and Figure 3 caption place the predicted transcriptional regulators in Figure 3. The methods say those regulators and their targets are shown in Figure 2.

  • Protein-array donor pooling differs

    Minor

    The results and figure caption describe two pooled samples containing three and four donors, respectively. The methods instead specify four donors per sample.

  • CD45+ comparison cites two different Zhang references

    Minor

    The figure caption attributes the CD45+ comparison to Zhang et al. reference 6, while the results attribute that comparison to reference 15. The paper's reference list identifies these as different Zhang publications.

5 of 5

Cites

Flagged first, then by citations

Cited by

Sorted by citations

Related

Authors
Galatro, T. F., Holtman, I. R., Lerario, A. M., Vainchtein, I. D., Brouwer, N., Sola, P. R., Veras, M. M., Pereira, T. F., Leite, R. E. P., Möller, T., Wes, P. D., Sogayar, M. C., Laman, J. D., den Dunnen, W., Pasqualucci, C. A., Oba-Shinjo, S. M., Boddeke, E. W. G. M., Marie, S. K. N., & Eggen, B. J. L.
Affiliations
  • University Medical Center Groningen (NL)
  • Universidade de São Paulo (BR)
  • Lundbeck (United States) (US)
Corresponding author
Eggen, Bart J L
ORCID
Venue
Nature Neuroscience
Publisher
Springer Science and Business Media LLC
Published
2017-07-03
Volume
20
Issue
8
Pages
1162-1171
ISSN
  • 1097-6256 (ISSN-L)
  • 1546-1726
DOI
10.1038/nn.4597
OpenAlex
W2729351500
Cited by
741
References
62
Citations by year
2026: 65 · 2025: 58 · 2024: 99 · 2023: 65 · 2022: 108 · 2021: 105 · 2020: 92 · 2019: 71 · 2018: 65 · 2017: 11 · 2012: 1
FWCI
32.12
Type
journal article
Language
en
Open access
Licence
Locations
  • Nature Neuroscience (publishedVersion)
  • PubMed (publishedVersion)
  • University of Groningen research database (University of Groningen / Centre for Information Technology) (publishedVersion, other-oa)
  • University of Groningen research database (University of Groningen / Centre for Information Technology) (submittedVersion, other-oa)
Topics
  • Life Sciences › Neuroscience › Neurology › Neuroinflammation and Neurodegeneration Mechanisms
  • Health Sciences › Medicine › Pathology and Forensic Medicine › Multiple Sclerosis Research Studies
  • Life Sciences › Biochemistry, Genetics and Molecular Biology › Molecular Biology › Single-cell and spatial transcriptomics
Keywords
microglia, cortical microglia, microglial gene expression, brain aging, human-mouse comparison, age-related gene expression, Fc gamma receptor, Siglec receptors, IFI16, TAL1, CX3CR1, P2Y12 receptor, CD11b, cell adhesion, actin cytoskeleton, neurodegeneration, parietal cortex
MeSH terms
Aging, Axons, Brain, Cell Cycle, Humans, Gene Expression, Microglia, Gene Expression Profiling, CD11b Antigen

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.

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