Skip to main content

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

  • Analysis
  • Published:

Cell cycle gene alterations associate with a redistribution of mutation risk across chromosomal domains in human cancers

Nature Cancer volume 5, pages 330–346 (2024)Cite this article

Subjects

Abstract

Mutations in human cells exhibit increased burden in heterochromatic, late DNA replication time (RT) chromosomal domains, with variation in mutation rates between tissues mirroring variation in heterochromatin and RT. We observed that regional mutation risk further varies between individual tumors in a manner independent of cell type, identifying three signatures of domain-scale mutagenesis in >4,000 tumor genomes. The major signature reflects remodeling of heterochromatin and of the RT program domains seen across tumors, tissues and cultured cells, and is robustly linked with higher expression of cell proliferation genes. Regional mutagenesis is associated with loss of activity of the tumor-suppressor genes RB1 and TP53, consistent with their roles in cell cycle control, with distinct mutational patterns generated by the two genes. Loss of regional heterogeneity in mutagenesis is associated with deficiencies in various DNA repair pathways. These mutation risk redistribution processes modify the mutation supply towards important genes, diverting the course of somatic evolution.

This is a preview of subscription content, access via your institution

Access options

Buy this article

  • Purchase on Springer Link
  • Instant access to full article PDF
Buy now

Prices may be subject to local taxes which are calculated during checkout

Fig. 1: Identifying RMD signatures by an NMF-based method applied to megabase-scale mutation density profiles.
Fig. 2: RMDglobal1 mutation risk signature associates with domain-scale variability in heterochromatin.
Fig. 3: RMDglobal1 mutation risk redistribution is linked with DNA replication time program changes in cancers.
Fig. 4: Genetic alterations in cancer genes are associated with the activity of RMDglobal1 mutagenesis.
Fig. 5: TP53 loss of function alterations underlie the RMDglobal2 mutation risk redistribution signature.

Similar content being viewed by others

Data availability

In this study, published datasets were reanalyzed. WGS somatic mutation calls for the PCAWG study were downloaded from the International Cancer Genome Consortium Data portal (https://dcc.icgc.org/pcawg). Restricted-access WGS somatic mutation calls for the HMF project were accessed via request number DR-260 (details at https://www.hartwigmedicalfoundation.nl/en/). WGS somatic mutation calls for the POG project were downloaded from BC Cancer (https://www.bcgsc.ca/downloads/POG570/). We downloaded restricted-access bam files for the TCGA (dbGaP accession phs000178.v11.p8), CPTAC (phs001287.v17.p6) and MMRF COMMPASS (phs000748.v7.p4) projects from the Genomic Data Commons data portal (https://portal.gdc.cancer.gov/).

We downloaded from ENCODE (https://www.encodeproject.org/) all data available for Homo sapiens in the genome assembly hg19 for DHS, H3F3A, H3K27me3, H3K4me1, H3K4me3, H3K9ac, H3K9me3, Hi-C, DNA methylation (WGBS), H2AFZ, H3K27ac, H3K36me3, H3K4me2, H3K79me2, H3K9me2 and H4K20me1 marks (described in Supplementary Table 7). We downloaded experimental RT data from the Replication Domain database (https://www2.replicationdomain.com/index.php). We downloaded ATAC-seq data of TCGA tumors (https://pubmed.ncbi.nlm.nih.gov/30361341/). We downloaded the chromatin domain hierarchies and compartment scores generated by the Calder method from Hi-C data from 114 cell lines (https://pubmed.ncbi.nlm.nih.gov/33972523/). We downloaded the 25 ChromHMM states segmented files (‘imputed12marks_segments’) for the 129 cell types available from Roadmap epigenomics (http://compbio.mit.edu/ChromHMM/).

Additionally, we downloaded other epigenomic data from various studies. The replication timing heterogeneity calculated as Twidth and Trep from high-resolution (16 phases) RepliSeq data were from ref. 73. The RT changes under overexpression of the oncogene KDM4A were from ref. 75. Five RT signatures of replication stress were from ref. 76. Ten RT cell type-specific signatures during development were from ref. 77. Fifteen RT states were from ref. 78. The changes (late to early or retain late) in RT upon RIF1 KO were from ref. 79. The RT changes due to RT QTLs were from ref. 80. The differences in RT between a hypomethylated cell line versus a control cell line were from ref. 19. The regions with variability in methylation across individuals were from ref. 81. The PMDs and HMDs were from ref. 20. The CpG density, gene density and LADs were from the Table Browser (https://genome.ucsc.edu/cgi-bin/hgTables) (assembly February 2009 GRCh37/hg19). The asynchronous replication domains were from ref. 82. The early-replicating fragile sites were from ref. 83. The SPIN states were from ref. 40. The A/B subcompartments were from ref. 39. Sixteen signatures generated from applying NMF to DHS peaks were from ref. 84. The H3K27me3 and H3K9me profiles for RB1 WT and KO were from ref. 23. The constitutive early, constitutive late and developmental domains were from http://www.replicationdomain.org.

In a FigShare repository (https://doi.org/10.6084/m9.figshare.c.6911140.v1), we provide data generated in this study: the RMD values across 2,450 1-Mb windows for the 4,221 tumor genomes analyzed (rmd_counts.zip) and the final RMD signatures extracted from this RMD matrix using NMF (RMDsignatures_exposures_k = 13_nFact13_n = 4221.csv and RMDsignatures_window_weights_hg19_k = 13_nFact=13.csv). In addition, we provide the RT and chromatin-remodeling PC signatures (PCA_chrom_RT.zip). Finally, we provide the predicted DNA replication timing data at 1-Mb resolution using the Replicon tool for TCGA tumors (predRT-TCGA_1Mb.zip) and for ENCODE samples (predRT-ENCODE_1Mb.zip). Other data can be made available from the authors upon request. Source data are provided with this paper.

Code availability

Custom code is available in a GitHub repository at https://github.com/marina-salvadores/RMDsig.

References

  1. Supek, F. & Lehner, B. Differential DNA mismatch repair underlies mutation rate variation across the human genome. Nature 521, 81–84 (2015).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

  2. Zheng, C. L. et al. Transcription restores DNA repair to heterochromatin, determining regional mutation rates in cancer genomes. Cell Rep. 9, 1228–1234 (2014).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  3. Pope, B. D. et al. Topologically associating domains are stable units of replication-timing regulation. Nature 515, 402–405 (2014).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

  4. Akdemir, K. C. et al. Somatic mutation distributions in cancer genomes vary with three-dimensional chromatin structure. Nat. Genet. 52, 1178–1188 (2020).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  5. Hansen, R. S. et al. Sequencing newly replicated DNA reveals widespread plasticity in human replication timing. Proc. Natl. Acad. Sci. USA 107, 139–144 (2010).

    Article  ADS  CAS  PubMed  Google Scholar 

  6. Polak, P. et al. Cell-of-origin chromatin organization shapes the mutational landscape of cancer. Nature 518, 360–364 (2015).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

  7. Makova, K. D. & Hardison, R. C. The effects of chromatin organization on variation in mutation rates in the genome. Nat. Rev. Genet. 16, 213–223 (2015).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  8. Schuster-Böckler, B. & Lehner, B. Chromatin organization is a major influence on regional mutation rates in human cancer cells. Nature 488, 504–507 (2012).

    Article  ADS  PubMed  Google Scholar 

  9. Salvadores, M., Mas-Ponte, D. & Supek, F. Passenger mutations accurately classify human tumors. PLoS Comput. Biol. 15, e1006953 (2019).

    Article  ADS  PubMed  PubMed Central  Google Scholar 

  10. Jiao, W. et al. A deep learning system accurately classifies primary and metastatic cancers using passenger mutation patterns. Nat. Commun. 11, 1–12 (2020).

    Article  Google Scholar 

  11. Kübler, K. et al. Tumor mutational landscape is a record of the pre-malignant state. Preprint at bioRxiv https://doi.org/10.1101/517565 (2019).

  12. Koren, A. et al. Genetic variation in human DNA replication timing. Cell 159, 1015–1026 (2014).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  13. McRae, A. F. et al. Identification of 55,000 replicated DNA methylation QTL. Sci. Rep. 8, 17605 (2018).

    Article  ADS  PubMed  PubMed Central  Google Scholar 

  14. Oliva, M. et al. DNA methylation QTL mapping across diverse human tissues provides molecular links between genetic variation and complex traits. Nat. Genet. 55, 112–122 (2023).

    Article  CAS  PubMed  Google Scholar 

  15. Gate, R. E. et al. Genetic determinants of co-accessible chromatin regions in activated T cells across humans. Nat. Genet. 50, 1140–1150 (2018).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  16. Kinker, G. S. et al. Pan-cancer single cell RNA-seq uncovers recurring programs of cellular heterogeneity. Nat. Genet. 52, 1208–1218 (2020).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  17. Barkley, D. et al. Cancer cell states recur across tumor types and form specific interactions with the tumor microenvironment. Nat. Genet. 54, 1192–1201 (2022).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  18. Du, Q. et al. Replication timing and epigenome remodelling are associated with the nature of chromosomal rearrangements in cancer. Nat. Commun. 10, 416 (2019).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

  19. Du, Q. et al. DNA methylation is required to maintain both DNA replication timing precision and 3D genome organization integrity. Cell Rep. 36, 109722 (2021).

    Article  CAS  PubMed  Google Scholar 

  20. Zhou, W. et al. DNA methylation loss in late-replicating domains is linked to mitotic cell division. Nat. Genet. 50, 591–602 (2018).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  21. Brinkman, A. B. et al. Partially methylated domains are hypervariable in breast cancer and fuel widespread CpG island hypermethylation. Nat. Commun. 10, 1749 (2019).

    Article  ADS  PubMed  PubMed Central  Google Scholar 

  22. Gurrion, C., Uriostegui, M. & Zurita, M. Heterochromatin reduction correlates with the increase of the KDM4B and KDM6A demethylases and the expression of pericentromeric DNA during the acquisition of a transformed phenotype. J. Cancer 8, 2866–2875 (2017).

    Article  PubMed  PubMed Central  Google Scholar 

  23. Wong, K. M., King, D. A., Schwartz, E. K., Herrera, R. E. & Morrison, A. J. Retinoblastoma protein regulates carcinogen susceptibility at heterochromatic cancer driver loci. Life Sci. Alliance 5, e202101134 (2022).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  24. Huang, Y., Gu, L. & Li, G.-M. H3K36me3-mediated mismatch repair preferentially protects actively transcribed genes from mutation. J. Biol. Chem. 293, 7811–7823 (2018).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  25. Poetsch, A. R., Boulton, S. J. & Luscombe, N. M. Genomic landscape of oxidative DNA damage and repair reveals regioselective protection from mutagenesis. Genome Biol. 19, 215 (2018).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  26. Alexandrov, L. B. et al. Signatures of mutational processes in human cancer. Nature 500, 415–421 (2013).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  27. Supek, F. & Lehner, B. Clustered mutation signatures reveal that error-prone DNA repair targets mutations to active genes. Cell 170, 534–547 (2017).

    Article  CAS  PubMed  Google Scholar 

  28. Degasperi, A. et al. Substitution mutational signatures in whole-genome–sequenced cancers in the UK population. Science 376, abl9283 (2022).

    Article  Google Scholar 

  29. Hoadley, K. A. et al. Multiplatform analysis of 12 cancer types reveals molecular classification within and across tissues of origin. Cell 158, 929–944 (2014).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  30. Mas-Ponte, D. & Supek, F. DNA mismatch repair promotes APOBEC3-mediated diffuse hypermutation in human cancers. Nat. Genet. 52, 958–968 (2020).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  31. Nguyen, L., W. M. Martens, J., Van Hoeck, A. & Cuppen, E. Pan-cancer landscape of homologous recombination deficiency. Nat. Commun. 11, 5584 (2020).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

  32. Yaacov, A. et al. Cancer mutational processes vary in their association with replication timing and chromatin accessibility. Cancer Res. 81, 6106–6116 (2021).

    Article  CAS  PubMed  Google Scholar 

  33. Supek, F. & Lehner, B. Scales and mechanisms of somatic mutation rate variation across the human genome. DNA Repair 81, 102647 (2019).

    Article  CAS  PubMed  Google Scholar 

  34. Liu, Y. et al. Systematic inference and comparison of multi-scale chromatin sub-compartments connects spatial organization to cell phenotypes. Nat. Commun. 12, 2439 (2021).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

  35. Liberzon, A. et al. The Molecular Signatures Database (MSigDB) hallmark gene set collection. Cell Syst. 1, 417 (2015).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  36. Moore, J. E. et al. Expanded encyclopaedias of DNA elements in the human and mouse genomes. Nature 583, 699–710 (2020).

    Article  ADS  PubMed  PubMed Central  Google Scholar 

  37. Corces, M. R. et al. The chromatin accessibility landscape of primary human cancers. Science 362, eaav1898 (2018).

    Article  ADS  PubMed  PubMed Central  Google Scholar 

  38. Gindin, Y., Meltzer, P. S. & Bilke, S. Replicon: a software to accurately predict DNA replication timing in metazoan cells. Front. Genet. 5, 378 (2014).

    Article  PubMed  PubMed Central  Google Scholar 

  39. Pratto, F. et al. Meiotic recombination mirrors patterns of germline replication in mice and humans. Cell 184, 4251–4267 (2021).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  40. Gnan, S. et al. Kronos scRT: a uniform framework for single-cell replication timing analysis. Nat. Commun. 13, 2329 (2022).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

  41. Rao, S. S. P. et al. A 3D Map of the Human Genome at Kilobase Resolution Reveals Principles of Chromatin Looping. Cell 159, 1665–1680 (2014).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  42. Wang, Y. et al. SPIN reveals genome-wide landscape of nuclear compartmentalization. Genome Biol. 22, 36 (2021).

    Article  PubMed  PubMed Central  Google Scholar 

  43. Lambuta, R. A. et al. Whole-genome doubling drives oncogenic loss of chromatin segregation. Nature 615, 925–933 (2023).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

  44. Gonzalo, S. et al. Role of the RB1 family in stabilizing histone methylation at constitutive heterochromatin. Nat. Cell Biol. 7, 420–428 (2005).

    Article  CAS  PubMed  Google Scholar 

  45. Krishnan, B. et al. Active RB causes visible changes in nuclear organization. J. Cell Biol. 221, e202102144 (2022).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  46. Dick, F. A., Goodrich, D. W., Sage, J. & Dyson, N. J. Non-canonical functions of the RB protein in cancer. Nat. Rev. Cancer 18, 442–451 (2018).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  47. Takahashi, C., Contreras, B., Bronson, R. T., Loda, M. & Ewen, M. E. Genetic Interaction between Rb and K-Ras in the control of differentiation and tumor suppression. Mol. Cell. Biol. 24, 10406–10415 (2004).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  48. Lee, K. Y., Ladha, M. H., McMahon, C. & Ewen, M. E. The retinoblastoma protein is linked to the activation of Ras. Mol. Cell. Biol. 19, 7724–7732 (1999).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  49. Dietlein, F. et al. Identification of cancer driver genes based on nucleotide context. Nat. Genet. 52, 208–218 (2020).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  50. Yaeger, R. et al. Clinical sequencing defines the genomic landscape of metastatic colorectal cancer. Cancer Cell 33, 125–136 (2018).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  51. Cramer, D., Serrano, L. & Schaefer, M. H. A network of epigenetic modifiers and DNA repair genes controls tissue-specific copy number alteration preference. eLife 5, e16519 (2016).

    Article  PubMed  PubMed Central  Google Scholar 

  52. Donehower, L. A. et al. Integrated analysis of TP53 gene and pathway alterations in the cancer genome atlas. Cell Rep. 28, 1370–1384 (2019).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  53. Stamatoyannopoulos, J. A. et al. Human mutation rate associated with DNA replication timing. Nat. Genet. 41, 393–395 (2009).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  54. Hodgkinson, A., Chen, Y. & Eyre-Walker, A. The large-scale distribution of somatic mutations in cancer genomes. Hum. Mutat. 33, 136–143 (2012).

    Article  CAS  PubMed  Google Scholar 

  55. Woo, Y. H. & Li, W.-H. DNA replication timing and selection shape the landscape of nucleotide variation in cancer genomes. Nat. Commun. 3, 1004 (2012).

    Article  ADS  PubMed  Google Scholar 

  56. Liu, L., De, S. & Michor, F. DNA replication timing and higher-order nuclear organization determine single-nucleotide substitution patterns in cancer genomes. Nat. Commun. 4, 1502 (2013).

    Article  ADS  PubMed  Google Scholar 

  57. Polak, P. et al. Reduced local mutation density in regulatory DNA of cancer genomes is linked to DNA repair. Nat. Biotechnol. 32, 71–75 (2014).

    Article  CAS  PubMed  Google Scholar 

  58. Morganella, S. et al. The topography of mutational processes in breast cancer genomes. Nat. Commun. 7, 11383 (2016).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

  59. García-Nieto, P. E. et al. Carcinogen susceptibility is regulated by genome architecture and predicts cancer mutagenesis. EMBO J. 36, 2829–2843 (2017).

    Article  PubMed  PubMed Central  Google Scholar 

  60. Hudson, T. J. et al. International network of cancer genome projects. Nature 464, 993–998 (2010).

    Article  ADS  CAS  PubMed  Google Scholar 

  61. Priestley, P. et al. Pan-cancer whole-genome analyses of metastatic solid tumours. Nature 575, 210–216 (2019).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

  62. Pleasance, E. et al. Pan-cancer analysis of advanced patient tumors reveals interactions between therapy and genomic landscapes. Nat. Cancer 1, 452–468 (2020).

    Article  CAS  PubMed  Google Scholar 

  63. Ellis, M. J. et al. Connecting genomic alterations to cancer biology with proteomics: the NCI Clinical Proteomic Tumor Analysis Consortium. Cancer Discov. 3, 1108–1112 (2013).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  64. Edwards, N. J. et al. The CPTAC Data Portal: a resource for cancer proteomics research. J. Proteome Res. 14, 2707–2713 (2015).

    Article  CAS  PubMed  Google Scholar 

  65. Walker, B. A. et al. A high-risk, double-hit, group of newly diagnosed myeloma identified by genomic analysis. Leukemia 33, 159–170 (2019).

    Article  CAS  PubMed  Google Scholar 

  66. Kim, S. et al. Strelka2: fast and accurate calling of germline and somatic variants. Nat. Methods 15, 591–594 (2018).

    Article  CAS  PubMed  Google Scholar 

  67. Salvadores, M., Fuster-Tormo, F. & Supek, F. Matching cell lines with cancer type and subtype of origin via mutational, epigenomic, and transcriptomic patterns. Sci. Adv. 6, eaba1862 (2020).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

  68. Derrien, T. et al. Fast computation and applications of genome mappability. PLoS ONE 7, e30377 (2012).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

  69. Ormond, C., Ryan, N. M., Corvin, A. & Heron, E. A. Converting single nucleotide variants between genome builds: from cautionary tale to solution. Brief. Bioinform. 22, bbab069 (2021).

    Article  PubMed  PubMed Central  Google Scholar 

  70. Amemiya, H. M., Kundaje, A. & Boyle, A. P. The ENCODE blacklist: identification of problematic regions of the genome. Sci. Rep. 9, 9354 (2019).

    Article  ADS  PubMed  PubMed Central  Google Scholar 

  71. Buisson, R. et al. Passenger hotspot mutations in cancer driven by APOBEC3A and mesoscale genomic features. Science 364, eaaw2872 (2019).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

  72. Tate, J. G. et al. COSMIC: the catalogue of somatic mutations in cancer. Nucleic Acids Res. 47, D941–D947 (2019).

    Article  CAS  PubMed  Google Scholar 

  73. Vali-Pour, M., Lehner, B. & Supek, F. The impact of rare germline variants on human somatic mutation processes. Nat. Commun. 13, 3724 (2022).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

  74. Kundaje, A. et al. Integrative analysis of 111 reference human epigenomes. Nature 518, 317–330 (2015).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  75. Zhao, P. A., Sasaki, T. & Gilbert, D. M. High-resolution Repli-Seq defines the temporal choreography of initiation, elongation and termination of replication in mammalian cells. Genome Biol. 21, 76 (2020).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  76. Sima, J. et al. Identifying cis elements for spatiotemporal control of mammalian DNA replication. Cell 176, 816–830 (2019).

    Article  CAS  PubMed  Google Scholar 

  77. Van Rechem, C. et al. Collective regulation of chromatin modifications predicts replication timing during cell cycle. Cell Rep. 37, 109799 (2021).

    Article  PubMed  PubMed Central  Google Scholar 

  78. Sarni, D. et al. 3D genome organization contributes to genome instability at fragile sites. Nat. Commun. 11, 3613 (2020).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

  79. Rivera-Mulia, J. C. et al. Dynamic changes in replication timing and gene expression during lineage specification of human pluripotent stem cells. Genome Res. 25, 1091–1103 (2015).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  80. Poulet, A. et al. RT states: systematic annotation of the human genome using cell type-specific replication timing programs. Bioinformatics 35, 2167–2176 (2019).

    Article  CAS  PubMed  Google Scholar 

  81. Klein, K. N. et al. Replication timing maintains the global epigenetic state in human cells. Science 372, 371–378 (2021).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

  82. Ding, Q. et al. The genetic architecture of DNA replication timing in human pluripotent stem cells. Nat. Commun. 12, 6746 (2021).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

  83. Gunasekara, C. J. et al. A genomic atlas of systemic interindividual epigenetic variation in humans. Genome Biol. 20, 105–105 (2019).

    Article  PubMed  PubMed Central  Google Scholar 

  84. Mukhopadhyay, R. et al. Allele-specific genome-wide profiling in human primary erythroblasts reveal replication program organization. PLoS Genet. 10, e1004319 (2014).

    Article  PubMed  PubMed Central  Google Scholar 

  85. Barlow, J. H. et al. Identification of early replicating fragile sites that contribute to genome instability. Cell 152, 620–632 (2013).

    Article  CAS  PubMed  PubMed Central  Google Scholar 

  86. Meuleman, W. et al. Index and biological spectrum of human DNase I hypersensitive sites. Nature 584, 244–251 (2020).

    Article  ADS  CAS  PubMed  PubMed Central  Google Scholar 

Download references

Acknowledgements

Work was supported by funding from an FPU fellowship of the Spanish government, Ministry of Universities to M.S., an ERC StG ‘HYPER-INSIGHT’ (757700) to F.S., Horizon2020 project ‘DECIDER’ (965193) to F.S., Spanish government project ‘REPAIRSCAPE’ (PID2020-118795GB-I00) to F.S., CaixaResearch project ‘POTENT-IMMUNO’ (HR22-00402) to F.S., an ICREA professorship to F.S., the SGR funding of the Catalan government (SGR 00616) to F.S., and the Severo Ochoa centers of excellence award of the Spanish government to the hosting institution IRB Barcelona.

This publication and the underlying research are partly facilitated by the HMF and the Center for Personalized Cancer Treatment, which have generated, analyzed and made available data for this research. In addition, data used in this publication were generated by the CPTAC (NCI/NIH). We acknowledge that the results published here are in part based upon data generated by the TCGA Research Network at http://cancergenome.nih.gov/.

We are grateful to Daniel Naro for retrieving and processing (alignment and variant calling) the WGS data from the CPTAC-3 and COMMPASS studies.

Author information

Authors and Affiliations

  1. Genome Data Science, Institute for Research in Biomedicine (IRB Barcelona), Barcelona Institute of Science and Technology, Barcelona, Spain

    Marina Salvadores & Fran Supek

  2. Catalan Institution for Research and Advanced Studies (ICREA), Barcelona, Spain

    Fran Supek

Authors
  1. Marina Salvadores
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Fran Supek
    View author publications

    You can also search for this author in PubMed Google Scholar

Contributions

M.S. performed data curation, wrote the code, performed all analyses and visualized results. F.S. and M.S. jointly devised the analyses, interpreted results and drafted the manuscript. F.S. conceived and supervised the study.

Corresponding author

Correspondence to Fran Supek.

Ethics declarations

Competing interests

The authors declare no competing interests.

Peer review

Peer review information

Nature Cancer thanks Subhajyoti De, Jan Korbel and the other, anonymous, reviewer(s) for their contribution to the peer review of this work. Primary Handling Editor:

Additional information

Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Extended data

Extended Data Fig. 1 RMD variability across tissues and individuals.

a) Variance explained for the first 25 PCs of a PCA on the RMD matrix (4221 samples x 2540 one-megabase windows), and a baseline (by the broken stick heuristic). b) Inter-individual variability (SD, standard deviation) within each tissue (n = 22) for the first PCs. c) PC3 and 4 separate various cancer types. d) Log likelihood of the clustering of the first 22 PCs using different numbers of k clusters for two cases (including and excluding PC1). e) Clustering into 18 clusters, showing the number of tumor samples from each cancer type that are assigned to each RMD-based cluster (Methods). f) Cluster assignment for breast cancer samples of triple negative (TN) subtype (n = 74) and samples with high APOBEC (>25% of mutations are in APOBEC contexts, n = 231). g) As controls, the PC1 separates colon MSI versus MSS tumors (n = 738), and PC7 separates lymphoid tumors (n = 88) according to their level of the SHM-associated mutational signature SBS9. h) PC7 window weights for chromosome 22 agree with the known SHM region.

Source data

Extended Data Fig. 2 Benchmarking of the NMF methodology using simulated cancer genomes.

a) Profiles for the 9 simulated signatures generated (first 100 windows of chr1 shown). The value is the increase of mutations it will generate when multiplied by the cancer type vector (x2, x3 or x5). In various scenarios considered, the number of windows affected by the increase can be 10%, 20% or 50%, as stated in the panel title: sig_[windows affected %]_[increase X]. b) Comparison of the number of signatures recovered by NMF depending on the 9 simulation scenarios (columns) and the characteristics of the 9 simulated signatures (rows). We can see which simulated signatures are more often recovered (with a circle) and how strong (color and size of the circle) and under which conditions. c) Testing whether 3 mutations / Mb is sufficient for extracting signatures. A PCA in the original dataset is compared to a PCA from a subsampled dataset (mutations removed from tumor genomes until all genomes have 3 mutations / Mb). The spectra of the first 10 PCs correlate very highly between the original and reduced-mutation dataset, suggesting robust NMD signatures. d) PC2 and PC3 tumor sample exposures correlation between the original, and the reduced-mutation datasets. In addition, in both datasets PC4 window weights correlates similarly strongly with the RMDglobal1 signature. The blue lines represent the linear regression and the gray shades the 95% confidence intervals.

Source data

Extended Data Fig. 3 The activity levels of the NMF-derived RMD signatures across various human tumor types (n = 4221 samples).

Tumor sample ‘exposures’ distributed across cancer types for the 13 RMD signatures extracted. Marked in green are the tumor samples considered to have a high contribution from a particular signature. The threshold corresponds to the tumor sample exposure value for which 99% of the MSI samples are recovered in the RMDflat signature (a positive control for association of RMD signatures with a biological feature). Boxplots: the center line is the median, the box bounds the 25th and 75th percentiles and the whiskers the largest/smallest value within 1.5 times the interquartile range (IQR).

Source data

Extended Data Fig. 4 Characterization of the RMDflat signature, a loss of megabase-scale mutation rate heterogeneity.

a) Correlation between RMDflat signature NMF window weights (n = 2540 windows) and the DNA replication timing (RT) (here, average RepliSeq signal across 10 cell lines). b) RMDflat signature exposures (that is activities) for groups of tumor samples with various DNA repair failures: (i) MSI, microsatellite instable, indicating a MMR failure, n = 147; (ii) HRD, homologous recombination deficiency, split by the BRCA1-type or BRCA2-type or not otherwise specified, n = 55 BRCA1, 116 BRCA2 and 126 NOS, or (iii) high levels of APOBEC mutation signatures, n = 497, (iv) ERCC2 mutant, n = 4, and (v) the remainder of the tumors are in the ‘not tested’ or ‘MSS’ groups (latter means that MSI was tested but was negative), n = 817 NotTested and 2459 MSS. Boxplots: the center line is the median, the box bounds the 25th and 75th percentiles and the whiskers the largest/smallest value within 1.5 times the interquartile range (IQR). c) Percentage of tumor samples with flat mutation rate landscapes (RMDflat exposure>0.177, a threshold that recovers 95% of MSI samples, n = 1137 samples) belonging to each of the 5 listed DNA repair categories, broken down by cancer type. The percentage of RMDflat-high samples in each cancer type is indicated in the x-axis labels. d) Association between the RMDflat activities and the tumor correlation with the average replication timing. The redistribution effects of the RMDflat signature reflects a strong increase in relative mutation rates in early replicating, euchromatic regions, and consequently reduced correlation of domain mutation risk with RT. MSI n = 147; HRD n = 297, high APOBEC n = 497, ERCC2 mutant n = 4, and the rest n = 3276. e) Distribution of the log2 difference in the intronic mutation rates (a measure of mutation supply towards a gene locus) for 460 cancer gene loci, comparing between RMDflat-high tumors and RMDflat-low tumors, using the actual values (‘RMDflat’ histogram) and randomized values (‘RMDflat randomized’ histogram). f) Log2 relative intronic mutation density (normalized to flanking DNA in same chromosome arm, see panel d) for the RMDflat-high versus the RMDflat-low, for 5 example common driver genes with highest effect sizes in this test. The dots are cancer types: n = 20 RMDflat_high and 20 RMDflat_low per gene. Boxplots as defined in b. g) Mean RMD profile across the DNA repair groups, shown example for chr 1p. Vertical lines mark the position for three of the example genes from panel f. APOBEC n = 497, HRD n = 297, MSI n = 147, MSS n = 2459, Not tested n = 817. h) RMD association with RT for HR-deficient and HR proficient tumors. Mutations are stratified by mutation types. SBS13-like are T(C > G)N mutations (APOBEC mutagenesis), while SBS3_proxy are [CAG](C > T)[CTA] mutations (HR deficiency-associated mutations). i) The indel mutational spectra for various groupings of tumor by RMDflat signature. On the y-axis ‘repe’ = repetition unit. As controls, the MSI tumors (upper left panel) show very high numbers of deletions of sizes -2 to -6, in the ‘repe1’ category (repeats with repeat unit 1 nucleotide long that is homopolymer microsatellites) than the known MSS tumors (lower left panel). The MSS_RMDflat-high category of tumors, which are high in the RMDflat signature but not known to be MSI nor APOBEC-high nor HR-deficient, have an indel spectrum resembling that of the MSS tumors, arguing against undetected MMR failures. For the NotTested_RMDflat_high, we identified 6 samples that showed an MSI-like indel spectrum suggesting they should be labeled MSI.

Source data

Extended Data Fig. 5 Quality control for the mutation risk redistribution signatures.

a) Silhouette index (clustering quality score) of each RMD signature. The silhouette index (robustness of profile to noise introduced in NMF runs) is comparable to the other RMD signatures. b) Autocorrelation (similarity in weights of consecutive 1 Mb windows) values with lag = 1 (calculated for each chromosome separately) for each RMD signature. n = 22 chromosomes. Boxplots: the center line is the median, the box bounds the 25th and 75th percentiles and the whiskers the largest/smallest value within 1.5 times the interquartile range (IQR). c) Correlation of RMD-indel profiles with the 3 global RMD signatures, stratifying the tumor samples into bins (20 tumors per bin) by their RMD signature exposure (x axis; derived from SNV mutations, not indels). The slope from the regression is shown next to each regression line. In specific, we binned the tumor samples according to their RMD signature exposure levels, and calculated the indel mutation density profile across the 1 Mb windows for the pooled indels across tumor samples in each bin (pooling is to increase statistical power, as the numbers of indels are lower than of SNV mutations). Then, we correlated these RMD-indel-profiles with the profile of 1 Mb window weights of the matching RMD signature. d-f) Correlation of RMD-SVs profiles (inversions, deletions and duplications) with the 3 global RMD signatures, stratifying the tumor samples into 20 bins by their RMD signature exposure (x axis; derived from SNV mutations). The slope from the regression is shown next to each regression line. In specific, we binned the tumor samples according to their RMD signature exposure levels, and calculated the DNA rearrangement mutation density profile across the 1 Mb windows for the pooled DNA rearrangements across tumor samples in each bin (pooling is to increase statistical power, as the numbers of DNA rearrangements are lower than of SNV mutations). We count each DNA rearrangement as 1 count (1 event) by taking into account only the start position. Then, we correlated these RMD-SVs-profiles with the profile of 1 Mb window weights of the matching RMD signature. c-f) The colored lines represent the linear regressions for each RMD signature and the colored shades the corresponding 95% confidence intervals. g) Prediction of RMD signatures from the density of ChromHMM states in chromosomal domains. Adjusted R2 of a regression predicting RMD signatures window weights from the chromHMM states density (% of the 1 Mb window covered by a particular state) (x axis). Predictions were made using either the whole dataset of the densities (many samples used as input for regression), the mean density of the feature (average over all samples used as input for regression) or selecting the maximum adjusted R2 of calculating R2 in regression for each sample individually.

Source data

Extended Data Fig. 6 Variability in ENCODE gene expression, histone marks and predicted RT correlates with tumor RMDglobal1.

a) Gene set enrichment Analysis (GSEA) scores for the Recurrent Heterogeneity Pathways (RHPs), for an ordered list of genes, based on the correlation of each gene’s expression to the exposures of the 3 epigenomic PCs across ENCODE samples. For each epigenomic PC, we predicted the PCx sample exposures from the expression of each gene separately (for example regression formula: H3K9me3_PC3 ∼ geneA). We used the coefficient for the gene variable from the regression (that is the effect size) to order the genes, and applied GSEA on the gene list ordered by effect size. The GSEA positive values means that the genes from that gene set are overall higher expressed in the PCx-high samples compared to the PCx-low samples. b) H3K27me3_PC3 sample exposures stratified by the most relevant biosample types in ENCODE. **** is p < =0.0001 by two-sided Mann-Whitney test. n = 80 cancer CL; 56 other CL, 51 primary cell; 94 tissue. c) Example of gene expression (sqrt TPMs) for a subset of genes from E2F targets and MYC targets categories of proliferation-associated genes, comparing ENCODE samples with high and low H3K27me3_PC3 exposures. n = 11 PC2-high-10%; 12 PC2-low-10%. d-e) Epigenetic-PCs ENCODE sample exposures stratified by the tissue types for H3K9me3_PC3, n = 256 (d) and (-)DHS_PC3, n = 676 (e). b-e) Boxplots: the center line is the median, the box bounds the 25th and 75th percentiles and the whiskers the largest/smallest value within 1.5 times the interquartile range (IQR). f) High correlation between predicted RT from ENCODE DHS (x axis) and experimental RT data from RepliSeq in the same cell lines (y axis). Example RT profiles for DHS-predicted RT and RepliSeq RT shown for two matched cell lines in the right panel. g) PredRT-TCGA_PC3 and PredRT-TCGA_PC4 are tissue-specific, separating: breast versus kidney cancer and brain versus colon cancer, respectively. h) The tumor sample exposures of the PredRT-TCGA_PC5, a RT PC that correlates with RMDglobal1 mutagenesis, broken down by cancer type, showing that cancers from almost all tissue types overlap by the range of RMDglobal1 scores (with a possible exception of the very rare cancer ‘PCPG’ pheochromocytoma/paraganglioma). n = 410 RT samples + 386 replicates. Boxplots as defined in b-e. i) Regression coefficients from a regression predicting the RMDglobal1 mutagenesis profile across 1 Mb windows jointly from multiple relevant epigenomic PCs and RT PCs. j) Gene set enrichment Analysis (GSEA) scores for the Recurrent Heterogeneity Pathways (RHPs) (Kinker et al. 2020), for an ordered list of genes based on the correlation of gene expression with RT PCs, and separately of that, gene expression with RMDglobal1 mutation risk. For each PC, we predicted the PCx sample exposures from the gene expression of each gene separately (for example regression formula: predRT-TCGA_PC5 ∼ geneA). We used the coefficient for the gene variable from the regression (that is the effect size) to order the genes, and applied GSEA on the gene list ordered by effect size.

Source data

Extended Data Fig. 7 Associations between RMDglobal1-associated chromatin restructuring PCs and chromatin features and between RMDglobal1 signature and RB1 deletion.

The panels show the chromatin PC window weights (n = 2540 windows) across, from left to right, (i) different Hi-C nuclear subcompartments; (ii) different SPIN nuclear spatial positioning states; (iii) correlation with the CORES score for Hi-C changes during a whole-genome doubling (the solid lines represent the linear regression and the shades the 95% confidence intervals); and (iv) compared to distance to telomeres, shown separately for (a) H3K9me3_PC3, (b) H3K27me3_PC2, (c) DHS_PC3 and (d) predRT-TCGA_PC5. P-values from two-sided Mann-Whitney test (ns: p > 0.05, *: p <= 0.05, **: p <= 0.01, ***: p <= 0.001 and ****: p <= 0.0001). e) Differences in RMDglobal1 exposures between RB1 deletion (-1 or -2 state, n = 581) and wild-type (0, +1 or +2 state, n = 2199) in a pan-cancer analysis (left) or by cancer type (right). f) Differences in RMDglobal1 exposures between a monoallelic (-1, n = 501) and biallelic (-2, n = 65) RB1 deletion and the wild-type (0, n = 1714). g) Differences in RMDglobal1 exposures between RB1 deletion (n = 500), RB1 mutations (n = 242) and wild-type (n = 2043). Boxplots: the center line is the median, the box bounds the 25th and 75th percentiles and the whiskers the largest/smallest value within 1.5*IQR.

Source data

Extended Data Fig. 8 Association analyses between RMD mutation risk signatures with genetic alterations in tumors and the contribution of clonal versus subclonal mutations.

a) The qq plots for p values in association testing of three RMD global signatures with CNA in cancer genes and chromatin modifying genes (see Methods). b) The qq plots for p values in association testing of three RMD global signatures with deleterious mutations in cancer genes and chromatin modifying genes (see Methods). a-b) The solid red lines represent the linear regression and the gray shades the 95% confidence intervals. c) RMD profiles for skin and lung tumors grouping samples by RMDglobal1 high or low activities. Windows with lowest and highest RMDglobal1 changes marked with vertical lines. d) Differences in RMDglobal1 exposures between KRAS mutated and wild-type tumors in a pan-cancer analysis (left) or by cancer type (right); p value by Mann-Whitney test. n = 507 KRAS mut and n = 1290 wt. e) Distribution of fraction C > A mutations (proxy for tobacco smoking exposure) along the patients separated by their self-reported smoking status. We classified the lung cancer samples with missing data (‘N/A’ in the plot) according to the C > A fraction threshold of 0.12 (vertical line). f) Differences in RMDglobal1 exposures between KRAS mutated and KRAS wild-type lung cancers, after stratifying by putative smoking (C > A > 12%) and non-smoking (C > A < 12%) status, shows that KRAS mutation is not associated with RMDglobal1 in smokers, but might be in non-smokers (nonsignificant trend, few non-smoker cancers are KRAS mutant). In lung adeno (luad) n = 6 KRAS mut versus n = 93 wt in non-smoking and n = 83 KRAS mut versus n = 133 wt in smoking. In unclassified lung n = 1 KRAS mut versus n = 11 wt in non-smoking and n = 11 KRAS mut versus n = 30 wt in smoking. P-values from two-sided Mann-Whitney test. g) RMD signature predicted exposures for different cancer types comparing the exposure for clonal versus subclonal profiles in Hartwig. n = 5231 clonal samples and 4596 subclonal samples. h) RMD signature predicted exposures for different cancer types comparing the exposure for clonal versus subclonal mutation profiles in CPTAC. n = 621 clonal samples and 621 subclonal samples. P-values from two-sided Mann-Whitney test (ns: p > 0.05, *: p <= 0.05, **: p <= 0.01, ***: p <= 0.001 and ****: p <= 0.0001). d-h) Boxplots: the center line is the median, the box bounds the 25th and 75th percentiles and the whiskers the largest/smallest value within 1.5*IQR.

Source data

Extended Data Fig. 9 Association of RMDglobal2 activities with: (i) TP53 deficiency and TP53 deficiency phenocopies; and (ii) copy number variants burden.

a) the association of regional mutation rates with RT is altered by the activity of RMDglobal2 signature. Mean RMD values across 10 RT bins (late to early) for samples with RMDglobal2_high versus RMDglobal2_low, shown across tissues. For RMDglobal2_high n = 5 breast, 32 lower-GI, 5 prostate, 7 upper-GI. For RMDglobal2_low n = 8 breast, 22 lower-GI, 2 prostate, 60 upper-GI. b) Association of the activity of the RMDglobal2 mutation risk signature with various mechanisms of TP53 inactivation in different cancer types. RMDglobal2 exposures grouped by: wild-type TP53 pathway (wt, n = 297), TP53 with 1 mutation (TP53_mut, n = 919), TP53 with 1 deletion (TP53_del, n = 124), TP53 loss phenocopied via an amplification in the MDM2, MDM4 or PPM1D genes (TP53_pheno, n = 416), or TP53 inactivation with two hits of either of the previously mentioned mechanisms (TP53_2hit, n = 973). Boxplots: the center line is the median, the box bounds the 25th and 75th percentiles and the whiskers the largest/smallest value within 1.5 times the interquartile range (IQR). c) Correlation between global RMD signatures activity and the CNV burden. d) Correlation between global RMDsignature2 and the CNV burden, upon stratifying by TP53 functional status. The solid blue lines represent the linear regression and the gray shades the 95% confidence intervals.

Source data

Extended Data Fig. 10 The association of the activity of the global RMD mutation risk redistribution signatures with the COSMIC trinucleotide mutational signatures.

a) Associations between the 3 global RMD signature exposures and the trinucleotide (COSMIC) mutational signatures exposures. We included all 3 RMD signatures in a single regression when testing for associations, such as to have the associations of one RMD signature be adjusted for the other two RMD signatures. P-values from two-sided Z-test on the regression coefficient. b) Correlations, broken down by cancer type, between RMDglobal2 and the SBS1 exposure (top positive hit). c) Correlations between RMDglobal2 and SBS93 (top negative hit). d) Correlations between RMDglobal1 and SBS17b (top negative hit). e) Correlations between RMDglobal1 and SBS_denovo_2 (top positive hit). b-e) The solid blue lines represent the linear regression and the gray shades the 95% confidence intervals. The R is the Pearson correlation coefficient with its associated p value.

Source data

Supplementary information

Reporting Summary

Supplementary Tables 1–9

Supplementary Table 1. Simulated ground-truth RMD signatures to benchmark the NMF methodology. Robustness of extracted RMD signatures, estimated as the SI for the clustering of the extracted RMD signatures at different k (number k-medoids clusters) and nFact (number NMF factors). It shows the minimum clustering SI for a particular condition (varying the signature contribution to the tumor mutational burden and the number of samples affected, as specified in the type column). Supplementary Table 2. Testing ability to identify NMF signatures matching the ground-truth simulated signatures. Cosine similarity values between the ground-truth RMD signatures (columns) and the extracted RMD signatures using NMF (rows). We consider a signature to be recovered correctly when it matches exactly one ground-truth RMD signature with cosine similarity ≥0.75. Supplementary Table 3. Replication of the pan-cancer RMD signatures in individual cancer types. Agreement between pan-cancer and individual cancer type extracted RMD signatures. Correlation between the RMD signatures extracted in the pan-cancer NMF analysis (main results) and in NMF runs by cancer type (additional analysis). It suggests that the three global RMD signatures (RMDflat, RMDglobal1 and RMDglobal2) can be recovered also from many individual cancer types. Supplementary Table 4. Prediction score (R2) of tissue-specific RMDglobal1 PCs in colon, esophagus, lung and sarcoma using different subsets of H3K9me3 profiles, grouped by tissue (the column n provides the sample size). As a control we generated 100 subsets of samples from mixed tissues with the sample sample size and calculated the R2, we provide the mean and s.d. of the R2 across the 100 runs. Supplementary Table 5. Epigenomic assays from different studies with reported association with RT. We tested whether these reported features correlated with our RMDglobal1 signature window weights. Information provided in the table is: source, feature name, the correlation with RMDglobal1 (R2), the assembly and a short description. Supplementary Table 6. WGS datasets description: source, number of samples, process and additional information. Supplementary Table 7. Number of samples downloaded from each epigenomic dataset from ENCODE. Supplementary Table 8. Genome-wide analysis testing all the genes bearing point mutations in at least ten tumor samples (n = 17,219 genes). We tested whether there is an association between a genetic feature (presence = 1 or absence = 0 of a deleterious mutation in a gene) and the three global RMD signatures exposures. In the same regression we control by tissue and total mutation burden of each sample. In this table we provide the estimates and P values for RMDglobal1. Supplementary Table 9. Genome-wide analysis testing all the genes bearing point mutations in at least ten tumor samples (n = 17,219 genes). We tested whether there is an association between a genetic feature (presence = 1 or absence = 0 of a deleterious mutation in a gene) and the three global RMD signatures exposures. In the same regression we control by tissue and total mutation burden of each sample. In this table we provide the estimates and P values for RMDglobal2.

Source data

Source Data Fig. 1

Statistical source data.

Source Data Fig. 2

Statistical source data.

Source Data Fig. 3

Statistical source data.

Source Data Fig. 4

Statistical source data.

Source Data Fig. 5

Statistical source data.

Source Data Extended Data Fig. 1

Statistical source data.

Source Data Extended Data Fig. 2

Statistical source data.

Source Data Extended Data Fig. 3

Statistical source data.

Source Data Extended Data Fig. 4

Statistical source data.

Source Data Extended Data Fig. 5

Statistical source data.

Source Data Extended Data Fig. 6

Statistical source data.

Source Data Extended Data Fig. 7

Statistical source data.

Source Data Extended Data Fig. 8

Statistical source data.

Source Data Extended Data Fig. 9

Statistical source data.

Source Data Extended Data Fig. 10

Statistical source data.

Rights and permissions

Springer Nature or its licensor (e.g. a society or other partner) holds exclusive rights to this article under a publishing agreement with the author(s) or other rightsholder(s); author self-archiving of the accepted manuscript version of this article is solely governed by the terms of such publishing agreement and applicable law.

Reprints and permissions

About this article

Check for updates. Verify currency and authenticity via CrossMark

Cite this article

Salvadores, M., Supek, F. Cell cycle gene alterations associate with a redistribution of mutation risk across chromosomal domains in human cancers. Nat Cancer 5, 330–346 (2024). https://doi.org/10.1038/s43018-023-00707-8

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1038/s43018-023-00707-8

This article is cited by

Search

Advanced search

Quick links

Nature Briefing: Cancer

Sign up for the Nature Briefing: Cancer newsletter — what matters in cancer research, free to your inbox weekly.

Get what matters in cancer research, free to your inbox weekly. Sign up for Nature Briefing: Cancer