The evolutionary history of 2,658 cancers

Gerstung, Moritz; Jolly, Clemency; Leshchiner, Ignaty; Dentro, Stefan C.; González, Santiago; Mitchell, Thomas J.; Rubanova, Yulia; Anur, Pavana; Rosebrock, Daniel; Yu, Kaixan; Tarabichi, Maxime; Deshwar, Amit G.; Wintersinger, Jeff; Kleinheinz, Kortine; Vázquez-García, Ignacio; Haase, Kerstin; Sengupta, Subhajit; Macintyre, Geoff; Malikić, Salem; Donmez, Nilgun; Livitz, Dimitri; Cmero, Marek; Demeulemeester, Jonas; Schumacher, Steve; Fan, Yu; Yao, Xiaotong; Schlesner, Matthias; Boutros, Paul C.; Bowtell, David D.L.; Zhu, Hongtu; Getz, Gad; Imieliński, Marcin; Beroukhim, Rameen; Şahinalp, S. Cenk; Ji, Yuan; Peifer, Martin; Markowetz, Florian; Mustonen, Ville; Yuan, Ke; Wang, Wenyi; Morris, Quaid; Spellman, Paul T.; Wedge, David C.; Loo, Peter Van; Evolution, Pcawg

doi:10.1101/161562

Cited by 274 publications

(508 citation statements)

References 50 publications

Supporting

Mentioning

444

Contrasting

Unclassified

Order By: Relevance

“…Absolute copy number was called using the ASCAT algorithm 53 . Whole genome doubling (WGD) status were called using the criteria described previously 54 . Chromosome and chromosome arm level gains and losses were retrieved from 17 .…”

Section: Genomic Alterationsmentioning

confidence: 99%

Pan-cancer computational histopathology reveals mutations, tumor composition and prognosis

Jung

Torné

et al. 2019

Preprint

Self Cite

109

166

View full text Add to dashboard Cite

Key points• Pan-cancer computational histopathology analysis with deep learning extracts histopathological patterns and accurately discriminates 28 cancer and 14 normal tissue types • Computational histopathology predicts whole genome duplications, focal amplifications and deletions, as well as driver gene mutations • Wide-spread correlations with gene expression indicative of immune infiltration and proliferation • Prognostic information augments conventional grading and histopathology subtyping in the majority of cancers AbstractHere we use deep transfer learning to quantify histopathological patterns across 17,396 H&E stained histopathology image slides from 28 cancer types and correlate these with underlying genomic and transcriptomic data. Pan-cancer computational histopathology (PC-CHiP) classifies the tissue origin across organ sites and provides highly accurate, spatially resolved tumor and normal distinction within a given slide. The learned computational histopathological features correlate with a large range of recurrent genetic aberrations, including whole genome duplications (WGDs), arm-level copy number gains and losses, focal amplifications and deletions as well as driver gene mutations within a range of cancer types. WGDs can be predicted in 25/27 cancer types (mean AUC=0.79) including those that were not part of model training. Similarly, we observe associations with 25% of mRNA transcript levels, which enables to learn and localise histopathological patterns of molecularly defined cell types on each slide. Lastly, we find that computational histopathology provides prognostic information augmenting histopathological subtyping and grading in the majority of cancers assessed, which pinpoints prognostically relevant areas such as necrosis or infiltrating lymphocytes on each tumour section. Taken together, these findings highlight the large potential of PC-CHiP to discover new molecular and prognostic associations, which can augment diagnostic workflows and lay out a rationale for integrating molecular and histopathological data.

show abstract

Section: Genomic Alterationsmentioning

confidence: 99%

Pan-cancer computational histopathology reveals mutations, tumor composition and prognosis

Jung

Torné

et al. 2019

Preprint

Self Cite

109

166

View full text Add to dashboard Cite

show abstract

“…Individually analyzed genomic aberrations were chosen according to the following criteria: (1) known function as early driver events (19,37); (2) among the most frequently observed aberrations in breast cancer (4, 5); (3) significantly different incidence between intrinsic subtypes (χ 2 -test P < 0.05 (4, 5)); (4) non-overlap with other selected events (i.e. only one CNA located on 8q24).…”

Section: Genomic Analysesmentioning

confidence: 99%

Re-definition of claudin-low as a breast cancer phenotype

Fougner

Bergholtz

Norum

et al. 2019

Preprint

View full text Add to dashboard Cite

The claudin-low breast cancer subtype is defined by gene expression characteristics and encompasses a remarkably diverse range of breast tumors. Here, we investigate genomic, transcriptomic, and clinical features of claudin-low breast tumors. We show that claudin-low is not simply a subtype analogous to the intrinsic subtypes (basal-like, HER2-enriched, luminal A, luminal B and normal-like) as previously portrayed, but is a complex additional phenotype which may permeate breast tumors of various intrinsic subtypes. Claudin-low tumors were distinguished by low genomic instability, mutational burden and proliferation levels, and high levels of immune and stromal cell infiltration. In other aspects, claudin-low tumors reflected characteristics of their intrinsic subtype. Finally, we have developed an alternative method for identifying claudin-low tumors and thereby uncovered potential weaknesses in the established claudin-low classifier. In sum, these findings elucidate the heterogeneity in claudin-low breast tumors, and substantiate a re-definition of claudin-low as a cancer phenotype.Contact informationC.F. christian.fougner@rr-research.noH.B. helga.bergholtz@rr-research.noJ.H.N. jens.henrik.norum@rr-research.noT.S. therese.sorlie@rr-research.no

show abstract

“…Many analyses that use tumour sequencing data, including the characterisation of clonal architecture (Dentro et al, 2018;Jamal-Hanjani et al, 2017;Nik-Zainal et al, 2012;Williams et al, 2018), assessment of cancer drivers and mutational processes (McGranahan et al, 2015), and timing of mutational events (Gerstung et al, 2018;Mitchell et al, 2018) are reliant on an unbiased estimation of the proportion of cancer cells containing individual mutations. Such cancer cell fractions (CCFs) are computed using the fraction of aligned reads supporting the variant allele (VAF), the tumour copy number profile and the tumour sample purity (Dentro et al, 2017).…”

Section: Introductionmentioning

confidence: 99%

Correcting reference bias from the Illumina Isaac aligner enables analysis of cancer genomes

Chubb

Frangou

et al. 2019

Preprint

Self Cite

View full text Add to dashboard Cite

Estimating the fraction of cancer cells with individual somatic mutations is central to many analyses in cancer genomics, including characterisation of clonal architecture and timing of mutational events. Estimation of these cancer cell fractions (CCFs) is contingent on unbiased assessment of the fraction of reads supporting variant alleles (VAFs). We demonstrate that VAFs computed by the Illumina Isaac pipeline, used in many large-scale sequencing projects includingThe 100,000 Genomes Project, are biased by the preferential soft clipping of reads supporting non-reference alleles (semi-aligned reads). We show that these biased VAFs can have deleterious effects on downstream analyses reliant on unbiased CCF estimates. While Isaac bias can be corrected through realignment with alternative parameters, this is computationally intensive. We therefore developed FixVAF, a tool for removing bias introduced by soft clipping of semi-aligned reads, facilitating downstream analyses without the need for realignment. FixVAF is freely available at https://github.com/danchubb/FixVAF.

show abstract

The evolutionary history of 2,658 cancers

Cited by 274 publications

References 50 publications

Pan-cancer computational histopathology reveals mutations, tumor composition and prognosis

Pan-cancer computational histopathology reveals mutations, tumor composition and prognosis

Re-definition of claudin-low as a breast cancer phenotype

Correcting reference bias from the Illumina Isaac aligner enables analysis of cancer genomes

Contact Info

Product

Resources

About