CoChat
Sign Up
Cited by 12

The DOI matches its Crossref record; 5 of the 5 references it cites were found in a registry or an index.

Developmental hematopoietic stem cell variation explains clonal hematopoiesis later in life

Kreger, Jesse, Mooney, Jazlyn A., Shibata, Darryl, MacLean, Adam L.·2024·Nature Communications
DOI 10.1038/s41467-024-54711-2PMC PMC11599844

Published version of a preprint

Open this paper in CoChat — free account

Abstract

Clonal hematopoiesis becomes increasingly common with age, but its cause is enigmatic because driver mutations are often absent. Serial observations infer weak selection indicating variants are acquired much earlier in life with unexplained initial growth spurts. Here we use fluctuating CpG methylation as a lineage marker to track stem cell clonal dynamics of hematopoiesis. We show, via the shared prenatal circulation of monozygotic twins, that weak selection conferred by stem cell variation created before birth can reliably yield clonal hematopoiesis later in life. Theory indicates weak selection will lead to dominance given enough time and large enough population sizes. Human hematopoiesis satisfies both these conditions. Stochastic loss of weakly selected variants is naturally prevented by the expansion of stem cell lineages during development. The dominance of stem cell clones created before birth is supported by blood fluctuating CpG methylation patterns that exhibit low correlation between unrelated individuals but are highly correlated between many elderly monozygotic twins. Therefore, clonal hematopoiesis driven by weak selection in later life appears to reflect variation created before birth.

Overview

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

Integrity signals

Registry

Crossref · 10.1038/s41467-024-54711-2

Cites

5 of 5 references verified0 citation errors · 0 not checkable

Statistics

Figures

Fig. 1 Clonal dynamics of HSCs are characterized by methylation profiles over lifetime. A Cartoon depicting HSC population dynamics from embryo development throughout a human lifespan for two twins. Colors represent different HSC clones. Created in BioRender. Kreger, J. (2024) BioRender.com/m16o751. B β distributions of two twins over time, see circled numbers in panel A. Colors represent the different individuals. The Pearson correlation coefficient between individuals over time is also shown (right panel). C Pearson correlation coefficients for average methylation profiles between MZ twins. Dots represent individual comparisons and hexagons represent means of datasets. D Pearson correlation coefficients between DZ twins. E Pearson correlation coefficients between unrelated individuals. F Pearson correlation coefficients for all of the three comparisons (compilation of the data presented in C–E), along with lines of best fit and corresponding 90% confidence intervals. The lines of best fit and confidence intervals are calculated using the means of datasets (hexagons).
Fig. 1 Clonal dynamics of HSCs are characterized by methylation profiles over lifetime. A Cartoon depicting HSC population dynamics from embryo development throughout a human lifespan for two twins. Colors represent different HSC clones. Created in BioRender. Kreger, J. (2024) BioRender.com/m16o751. B β distributions of two twins over time, see circled numbers in panel A. Colors represent the different individuals. The Pearson correlation coefficient between individuals over time is also shown (right panel). C Pearson correlation coefficients for average methylation profiles between MZ twins. Dots represent individual comparisons and hexagons represent means of datasets. D Pearson correlation coefficients between DZ twins. E Pearson correlation coefficients between unrelated individuals. F Pearson correlation coefficients for all of the three comparisons (compilation of the data presented in C–E), along with lines of best fit and corresponding 90% confidence intervals. The lines of best fit and confidence intervals are calculated using the means of datasets (hexagons).CC BY · PubMed Central
Fig. 2 Variation during development is needed in order for model to match initial Pearson coefficients at birth. A Schematic diagram of the mathematical model, with probability of cell replacement α and probability of change in methylation γ. B HSC population dynamics over a lifetime. C Parameter estimation using initial Pearson coefficients of twins and unrelated individuals. Given a choice for the initial number of clones, the number of cells at which the embryo splits (Nsplit) can be determined. Data points (dots) represent the mean Pearson coefficient for 103 independent simulations of the model with a given number of clones (Nclones) and number of shared cells (Nsplit). Shaded areas and error bars denote one standard deviation from the mean over the 103 simulations. The horizontal lines represent the data: initial Pearson coefficients at birth for MZ twins (red), DZ twins (blue) and unrelated individuals (yellow). D–E Effect of varying Nsplit. White striped bars represent the clonal distribution in the shared embryo when it reaches Nsplit cells and splits into two embryos. Yellow/blue bars represent individuals 1 or 2 at the point at which the embryo has reached Ncells cells and finished growing. D: Nsplit = 10. E: Nsplit = 500. All other parameter values can be found in Table 2.
Fig. 2 Variation during development is needed in order for model to match initial Pearson coefficients at birth. A Schematic diagram of the mathematical model, with probability of cell replacement α and probability of change in methylation γ. B HSC population dynamics over a lifetime. C Parameter estimation using initial Pearson coefficients of twins and unrelated individuals. Given a choice for the initial number of clones, the number of cells at which the embryo splits (Nsplit) can be determined. Data points (dots) represent the mean Pearson coefficient for 103 independent simulations of the model with a given number of clones (Nclones) and number of shared cells (Nsplit). Shaded areas and error bars denote one standard deviation from the mean over the 103 simulations. The horizontal lines represent the data: initial Pearson coefficients at birth for MZ twins (red), DZ twins (blue) and unrelated individuals (yellow). D–E Effect of varying Nsplit. White striped bars represent the clonal distribution in the shared embryo when it reaches Nsplit cells and splits into two embryos. Yellow/blue bars represent individuals 1 or 2 at the point at which the embryo has reached Ncells cells and finished growing. D: Nsplit = 10. E: Nsplit = 500. All other parameter values can be found in Table 2.CC BY · PubMed Central
Fig. 3 Frequency-dependent growth produces clonal variation during development. A Uniform growth during development with Nsplit = 36 and no selection (si = 0) for all clones. Left panel: Clonal dynamics during development, colors denote different clones. Right panel: β distributions at the end of development, i.e. birth. The initial Pearson correlation coefficient at birth is 0.99. B Same as A for frequency-dependent growth model during development. The initial Pearson correlation coefficient at birth is 0.87. All other parameter values can be found in Table 2.
Fig. 3 Frequency-dependent growth produces clonal variation during development. A Uniform growth during development with Nsplit = 36 and no selection (si = 0) for all clones. Left panel: Clonal dynamics during development, colors denote different clones. Right panel: β distributions at the end of development, i.e. birth. The initial Pearson correlation coefficient at birth is 0.99. B Same as A for frequency-dependent growth model during development. The initial Pearson correlation coefficient at birth is 0.87. All other parameter values can be found in Table 2.CC BY · PubMed Central
Fig. 4 Variants with weak selection arise during development and explain FMC dynamics for monozygotic twins. Simulations shown for MZ twins (Nsplit = 36) with frequency-dependent growth during development. For plots of the data (middle column and bottom row): dots represent individual comparisons, hexagons represent means of datasets, and shaded bands represent 90% confidence intervals. The data plotted in each panel (A-C and D-F) are the same for purposes of comparison against different model simulations. A–C Clone growth frequency plots for both individuals during development and life (dashed vertical line represents Nsplit), Pearson correlation coefficient, and β distributions at 0, 50, and 100 years of life. A: No selection. B: Weak selection (a = 0.05 and θ = 0.01). C: Strong selection (a = 0.05 and θ = 0.05). D–F Results from 102 simulations are shown (dashed lines are individual simulations and solid lines are mean trajectories). D: No selection. E: Weak selection (a = 0.05 and θ = 0.01). F: Strong selection (a = 0.05 and θ = 0.05). All other parameter values can be found in Table 2.
Fig. 4 Variants with weak selection arise during development and explain FMC dynamics for monozygotic twins. Simulations shown for MZ twins (Nsplit = 36) with frequency-dependent growth during development. For plots of the data (middle column and bottom row): dots represent individual comparisons, hexagons represent means of datasets, and shaded bands represent 90% confidence intervals. The data plotted in each panel (A-C and D-F) are the same for purposes of comparison against different model simulations. A–C Clone growth frequency plots for both individuals during development and life (dashed vertical line represents Nsplit), Pearson correlation coefficient, and β distributions at 0, 50, and 100 years of life. A: No selection. B: Weak selection (a = 0.05 and θ = 0.01). C: Strong selection (a = 0.05 and θ = 0.05). D–F Results from 102 simulations are shown (dashed lines are individual simulations and solid lines are mean trajectories). D: No selection. E: Weak selection (a = 0.05 and θ = 0.01). F: Strong selection (a = 0.05 and θ = 0.05). All other parameter values can be found in Table 2.CC BY · PubMed Central
Fig. 5 Weak selection with variants arising in development explains lifetime dynamics of twins and unrelated individuals. Simulations shown for weak selection (a = 0.05 and θ = 0.01) with frequency-dependent growth during development. For plots of the data (middle column and bottom row): dots represent individual comparisons, hexagons represent means of datasets, and shaded bands represent 90% confidence intervals. The data plotted in each panel (A-C and D-F) are the same for purposes of comparison against different model simulations. A–C Clone growth frequency plots for both individuals during development and life (dashed vertical line represents Nsplit), Pearson correlation coefficient, and β distributions at 0, 50, and 100 years of life. A: MZ twins, Nsplit = 36. B: DZ twins, Nsplit = 15. C: Unrelated individuals, Nsplit = 10. D–F: Results from 102 simulations are shown (dashed lines are individual simulations and solid lines are mean trajectories). D: MZ twins. E: DZ twins. F: Unrelated individuals. All other parameter values can be found in Table 2.
Fig. 5 Weak selection with variants arising in development explains lifetime dynamics of twins and unrelated individuals. Simulations shown for weak selection (a = 0.05 and θ = 0.01) with frequency-dependent growth during development. For plots of the data (middle column and bottom row): dots represent individual comparisons, hexagons represent means of datasets, and shaded bands represent 90% confidence intervals. The data plotted in each panel (A-C and D-F) are the same for purposes of comparison against different model simulations. A–C Clone growth frequency plots for both individuals during development and life (dashed vertical line represents Nsplit), Pearson correlation coefficient, and β distributions at 0, 50, and 100 years of life. A: MZ twins, Nsplit = 36. B: DZ twins, Nsplit = 15. C: Unrelated individuals, Nsplit = 10. D–F: Results from 102 simulations are shown (dashed lines are individual simulations and solid lines are mean trajectories). D: MZ twins. E: DZ twins. F: Unrelated individuals. All other parameter values can be found in Table 2.CC BY · PubMed Central

Verification

Claims the paper makes

The claims and statistics of this paper haven’t been checked yet.

Run the check — free, no account needed. It reads the open-access text and checks the sources it cites.

A few free checks a day · papers already verified open instantly · results are public

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?

Self-check

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

The self-check needs the paper’s full text.

Citations

Cites

Show all 5 references

Research works printed but not linked: 72

  1. Haas, S., Trumpp, A. & Milsom, M. D. Causes and Consequences of Hematopoietic Stem Cell Heterogeneity. Cell Stem Cell 22, 627–638 (2018). · DOI
  2. Zink, F. et al. Clonal hematopoiesis, with and without candidate driver mutations, is common in the elderly. Blood 130, 742–752 (2017). · DOI · Paper page
  3. Robertson, N. A. et al. Longitudinal dynamics of clonal hematopoiesis identifies gene-specific fitness effects. Nat. Med. 28, 1439–1446 (2022). · DOI · Paper page
  4. Uddin, M. M. et al. Long-term longitudinal analysis of 4,187 participants reveals insights into determinants of clonal hematopoiesis. Nat. Commun. 15, 7858 (2024). · DOI
  5. Fabre, M. A. & Vassiliou, G. S. The lifelong natural history of clonal hematopoiesis and its links to myeloid neoplasia. Blood 143, 573–581 (2024). · DOI
  6. Ahmad, H., Jahn, N. & Jaiswal, S. Clonal Hematopoiesis and Its Impact on Human Health. Annu. Rev. Med. 74, 249–260 (2023). · DOI · Paper page
  7. MacLean, A. L., Lo Celso, C. & Stumpf, M. P. H. Concise review: stem cell population biology: insights from hematopoiesis. Stem Cells (Dayt., Ohio) 35, 80–88 (2017). · DOI
  8. Greaves, M. & Maley, C. C. Clonal evolution in cancer. Nature 481, 306–313 (2012). · DOI · Paper page
  9. Wu, W.-C. et al. Circulating hematopoietic stem and progenitor cells are myeloid-biased in cancer patients. Proc. Natl. Acad. Sci. 111, 4221–4226 (2014). · DOI
  10. Stadler, T., Pybus, O. G. & Stumpf, M. P. H. Phylodynamics for cell biologists. Science 371, eaah6266 (2021). · DOI
  11. Arends, C. M. & Jaiswal, S. Dynamics of clonal hematopoiesis and risk of hematologic malignancy. Int. J. Hematol. https://doi.org/10.1007/s12185-024-03829-6 (2024). · DOI
  12. Uddin, Md. M. et al. Clonal hematopoiesis of indeterminate potential, DNA methylation, and risk for coronary artery disease. Nat. Commun. 13, 5350 (2022). · DOI
  13. Mitchell, E. et al. Clonal dynamics of haematopoiesis across the human lifespan. Nature 606, 343–350 (2022). · DOI · Paper page
  14. Stacey, S. N. et al. Genetics and epidemiology of mutational barcode-defined clonal hematopoiesis. Nat. Genet. https://www.nature.com/articles/s41588-023-01555-z (2023).
  15. Weinstock, J. S. et al. Aberrant activation of TCL1A promotes stem cell expansion in clonal haematopoiesis. Nature 616, 755–763 (2023). · DOI · Paper page
  16. Kimura, M. On the probability of fixation of mutant genes in a population. Genetics 47, 713–719 (1962). · DOI
  17. Johnson, B., Shuai, Y., Schweinsberg, J. & Curtius, K. clonerate: fast estimation of single-cell clonal dynamics using coalescent theory. Bioinformatics 39, btad561 (2023). · DOI
  18. Fabre, M. A. et al. Concordance for clonal hematopoiesis is limited in elderly twins. Blood 135, 269–273 (2020). · DOI
  19. Ghersi, J. J. et al. Haematopoietic stem and progenitor cell heterogeneity is inherited from the embryonic endothelium. Nat. Cell Biol. 25, 1135–1145 (2023). · DOI · Paper page
  20. Wright, S. Evolution in Mendelian populations. Genetics 16, 97 (1931). · DOI
  21. Fisher, R. A. The evolution of dominance in certain polymorphic species. Am. Naturalist 64, 385–406 (1930). · DOI
  22. Moran, P. A. P. Random processes in genetics. Math. Proc. Camb. Philos. Soc. 54, 60–71 (1958). · DOI
  23. Levin, S. A. Dispersion and Population Interactions. Am. Naturalist 108, 207–228 (1974). · DOI
  24. Levin, S. A. & Powell, T. M.Patch Dynamics, vol. 96 of Lecture Notes in Biomathematics (Springer-Verlag Berlin Heidelberg, https://doi.org/10.1007/978-3-642-50155-5 1993). · DOI
  25. Otto, S. P. & Whitlock, M. C. The probability of fixation in populations of changing size. Genetics 146, 723–733 (1997). · DOI
  26. Patwa, Z. & Wahl, L. The fixation probability of beneficial mutations. J. R. Soc. Interface 5, 1279–1289 (2008). · DOI
  27. Nicholson, A. M. et al. Fixation and spread of somatic mutations in adult human colonic epithelium. Cell Stem Cell 22, 909–918.e8 (2018). · DOI
  28. Cosgrove, J., Hustin, L. S. P., de Boer, R. J. & Perié, L. Hematopoiesis in numbers. Trends Immunol. 42, 1100–1112 (2021). · DOI
  29. Marceau, K. et al. The prenatal environment in twin studies: a review on chorionicity. Behav. Genet. 46, 286–303 (2016). · DOI
  30. van Beijsterveldt, C. E. M. et al. Chorionicity and heritability estimates from twin studies: the prenatal environment of twins and their resemblance across a large number of traits. Behav. Genet. 46, 304–314 (2016). · DOI
  31. Hansen, J. W. et al. Clonal hematopoiesis in elderly twins: concordance, discordance, and mortality. Blood 135, 261–268 (2020). · DOI
  32. Trejo, V. et al. X Chromosome inactivation patterns correlate with fetal-placental anatomy in monozygotic twin pairs: implications for immune relatedness and concordance for autoimmunity. Mol. Med. 1, 62–70 (1994). · DOI
  33. Kristiansen, M. et al. Twin study of genetic and aging effects on X chromosome inactivation. Eur. J. Hum. Genet. 13, 599–606 (2005). · DOI
  34. Gabbutt, C. et al. Fluctuating methylation clocks for cell lineage tracing at high temporal resolution in human tissues. Nat. Biotechnol. 40, 720–730 (2022). · DOI · Paper page
  35. Müllers, S. M., McAuliffe, F. & Malone, F. D. 44 - Multiple Pregnancy. In Pandya, P. P., Oepkes, D., Sebire, N. J. & Wapner, R. J. (eds.) Fetal Medicine (Third Edition), 532–553.e6 (Elsevier, London, 2020). https://www.sciencedirect.com/science/article/pii/B9780702069567000440 · DOI
  36. Wenstrom, K. D., Tessen, J. A., Zlatnik, F. J. & Sipes, S. L. Frequency, distribution, and theoretical mechanisms of hematologic and weight discordance in monochorionic twins. Obstet. Gynecol. 80, 257 (1992).
  37. Park, S. et al. Clonal dynamics in early human embryogenesis inferred from somatic mutation. Nature 597, 393–397 (2021). · DOI
  38. Svensson, E. I. & Connallon, T. How frequency-dependent selection affects population fitness, maladaptation and evolutionary rescue. Evolut. Appl. 12, 1243–1258 (2019). · DOI
  39. Waxman, D. A unified treatment of the probability of fixation when population size and the strength of selection change over time. Genetics 188, 907–913 (2011). · DOI
  40. Lo Celso, C. & Scadden, D. T. The haematopoietic stem cell niche at a glance. J. Cell Sci. 124, 3529–3535 (2011). · DOI
  41. Paul, F. et al. Transcriptional heterogeneity and lineage commitment in myeloid progenitors. Cell 163, 1663–1677 (2015). · DOI · Paper page
  42. Rodriguez-Fraticelli, A. E. et al. Single-cell lineage tracing unveils a role for TCF15 in haematopoiesis. Nature 583, 585–589 (2020). · DOI · Paper page
  43. Ranzoni, A. M. et al. Integrative Single-Cell RNA-Seq and ATAC-Seq analysis of human developmental hematopoiesis. Cell Stem Cell 28, 472–487.e7 (2021). · DOI · Paper page
  44. Pei, W. et al. Resolving Fates and Single-Cell Transcriptomes of Hematopoietic Stem Cell Clones by PolyloxExpress Barcoding. Cell Stem Cell 27, 383–395.e8 (2020). · DOI · Paper page
  45. Sun, J. et al. Clonal dynamics of native haematopoiesis. Nature 514, 322–327 (2014). · DOI · Paper page
  46. Haltalli, M. L. R. et al. Manipulating niche composition limits damage to haematopoietic stem cells during Plasmodium infection. Nat. Cell Biol. 22, 1399–1410 (2020). · DOI
  47. Kessler, M. D. et al. Common and rare variant associations with clonal haematopoiesis phenotypes. Nature 612, 301–309 (2022). · DOI · Paper page
  48. Bao, E. L. et al. Inherited myeloproliferative neoplasm risk affects haematopoietic stem cells. Nature 586, 769–775 (2020). · DOI
  49. Brown, D. W. et al. Shared and distinct genetic etiologies for different types of clonal hematopoiesis. Nat. Commun. 14, 5536 (2023). · DOI
  50. Maley, C. C. et al. Classifying the evolutionary and ecological features of neoplasms. Nat. Rev. Cancer 17, 605–619 (2017). · DOI
  51. Weng, C. et al. Deciphering cell states and genealogies of human haematopoiesis. Nature 627, 389–398 (2024). · DOI · Paper page
  52. Watson, C. J. et al. Evolutionary dynamics in the decades preceding acute myeloid leukaemia. https://www.biorxiv.org/content/10.1101/2024.07.05.602251v1 (2024). · DOI
  53. Christie, M. R. & McNickle, G. G. Negative frequency dependent selection unites ecology and evolution. Ecol. Evolution 13, e10327 (2023). · DOI
  54. Barrett, T. et al. NCBI GEO: archive for functional genomics data sets-update. Nucleic Acids Res. 41, D991–D995 (2013). · DOI · Paper page
  55. K Mogensen, P. & N Riseth, A. Optim: A mathematical optimization package for Julia. J. Open Source Softw. 3, 615 (2018). · DOI
  56. Pearson, K. & Galton, F. VII. Note on regression and inheritance in the case of two parents. Proc. R. Soc. Lond. 58, 240–242 (1997).
  57. Cedar, H. & Bergman, Y. Programming of DNA methylation patterns. Annu. Rev. Biochem. 81, 97–117 (2012). · DOI
  58. Besançon, M. et al. Distributions.jl: Definition and modeling of probability distributions in the juliastats ecosystem. J. Stat. Softw. 98, 1–30 (2021). · DOI
  59. Kousathanas, A. & Keightley, P. D. A comparison of models to infer the distribution of fitness effects of new mutations. Genetics 193, 1197–1208 (2013). · DOI
  60. Kim, B. Y., Huber, C. D. & Lohmueller, K. E. Inference of the distribution of selection coefficients for new nonsynonymous mutations using large samples. Genetics 206, 345–361 (2017). · DOI
  61. Huber, C. D., Kim, B. Y., Marsden, C. D. & Lohmueller, K. E. Determining the factors driving selective effects of new nonsynonymous mutations. Proc. Natl. Acad. Sci. USA 114, 4465–4470 (2017). · DOI
  62. Bezanson, J., Edelman, A., Karpinski, S. & Shah, V. B. Julia: a fresh approach to numerical computing. SIAM Rev. 59, 65–98 (2017). · DOI · Paper page
  63. Roesch, E. et al. Julia for biologists. Nat. Methods 20, 655–664 (2023). · DOI
  64. Gordon, L. et al. Neonatal DNA methylation profile in human twins is specified by a complex interplay between intrauterine environmental and genetic factors, subject to tissue-specific influence. Genome Res. 22, 1395–1406 (2012). · DOI
  65. Hannon, E. et al. Characterizing genetic and environmental influences on variable DNA methylation using monozygotic and dizygotic twins. PLoS Genet. 14, e1007544 (2018). · DOI
  66. Ollikainen, M. et al. Genome-wide blood DNA methylation alterations at regulatory elements and heterochromatic regions in monozygotic twins discordant for obesity and liver fat. Clin. Epigenetics 7, 39 (2015). · DOI
  67. Tan, Q. et al. Epigenetic signature of birth weight discordance in adult twins. BMC Genomics 15, 1062 (2014). · DOI
  68. Roos, L. et al. Integrative DNA methylome analysis of pan-cancer biomarkers in cancer discordant monozygotic twin-pairs. Clin. Epigenetics 8, 7 (2016). · DOI
  69. Li, S. et al. Genetic and environmental causes of variation in the difference between biological age based on DNA methylation and chronological age for middle-aged women. Twin Res. Hum. Genet.: Off. J. Int. Soc. Twin Stud. 18, 720–726 (2015). · DOI
  70. Tan, Q. et al. Epigenetic drift in the aging genome: a ten-year follow-up in an elderly twin cohort. Int. J. Epidemiol. 45, 1146–1158 (2016).
  71. Ushijima, T. et al. Fidelity of the methylation pattern and its variation in the genome. Genome Res. 13, 868–874 (2003). · DOI
  72. Crofts, S. J. C., Latorre-Crespo, E. & Chandra, T. DNA methylation rates scale with maximum lifespan across mammals. https://www.biorxiv.org/content/10.1101/2023.05.15.540689v1 (2023). · DOI

Cited by

Metadata

Authors
Kreger, J., Mooney, J. A., Shibata, D., & MacLean, A. L.
Venue
Nature Communications
Publisher
Springer Science and Business Media LLC
Published
2024-11-26
Volume
15
Issue
1
ISSN
2041-1723
DOI
10.1038/s41467-024-54711-2
PMC
PMC11599844
Cited by
12
References
5
Language
en
Licence
Funders
U.S. Department of Health & Human Services | National Institutes of Health (R35GM143019)

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.

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