Skip to main content

Lund University Publications

LUND UNIVERSITY LIBRARIES

A deep learning system accurately classifies primary and metastatic cancers using passenger mutation patterns

Jiao, Wei ; Atwal, Gurnit ; Polak, Paz ; Karlic, Rosa ; Cuppen, Edwin ; Danyi, Alexandra ; de Ridder, Jeroen ; van Herpen, Carla ; Lolkema, Martijn P and Steeghs, Neeltje , et al. (2020) In Nature Communications 11.
Abstract

In cancer, the primary tumour's organ of origin and histopathology are the strongest determinants of its clinical behaviour, but in 3% of cases a patient presents with a metastatic tumour and no obvious primary. Here, as part of the ICGC/TCGA Pan-Cancer Analysis of Whole Genomes (PCAWG) Consortium, we train a deep learning classifier to predict cancer type based on patterns of somatic passenger mutations detected in whole genome sequencing (WGS) of 2606 tumours representing 24 common cancer types produced by the PCAWG Consortium. Our classifier achieves an accuracy of 91% on held-out tumor samples and 88% and 83% respectively on independent primary and metastatic samples, roughly double the accuracy of trained pathologists when... (More)

In cancer, the primary tumour's organ of origin and histopathology are the strongest determinants of its clinical behaviour, but in 3% of cases a patient presents with a metastatic tumour and no obvious primary. Here, as part of the ICGC/TCGA Pan-Cancer Analysis of Whole Genomes (PCAWG) Consortium, we train a deep learning classifier to predict cancer type based on patterns of somatic passenger mutations detected in whole genome sequencing (WGS) of 2606 tumours representing 24 common cancer types produced by the PCAWG Consortium. Our classifier achieves an accuracy of 91% on held-out tumor samples and 88% and 83% respectively on independent primary and metastatic samples, roughly double the accuracy of trained pathologists when presented with a metastatic tumour without knowledge of the primary. Surprisingly, adding information on driver mutations reduced accuracy. Our results have clinical applicability, underscore how patterns of somatic passenger mutations encode the state of the cell of origin, and can inform future strategies to detect the source of circulating tumour DNA.

(Less)
Please use this url to cite or link to this publication:
author
; ; ; ; ; ; ; ; and , et al. (More)
; ; ; ; ; ; ; ; ; ; ; ; ; and (Less)
contributor
LU ; LU orcid and LU orcid
author collaboration
organization
publishing date
type
Contribution to journal
publication status
published
subject
keywords
Computational Biology/methods, Deep Learning, Female, Genome, Human, Humans, Male, Mutation, Neoplasm Metastasis, Neoplasms/genetics, Reproducibility of Results, Whole Genome Sequencing
in
Nature Communications
volume
11
article number
728
pages
12 pages
publisher
Nature Publishing Group
external identifiers
  • pmid:32024849
  • scopus:85079062177
ISSN
2041-1723
DOI
10.1038/s41467-019-13825-8
language
English
LU publication?
yes
id
da9f791d-9dc3-4b4a-abe2-5950389be426
date added to LUP
2023-01-05 14:46:27
date last changed
2025-07-12 04:34:24
@article{da9f791d-9dc3-4b4a-abe2-5950389be426,
  abstract     = {{<p>In cancer, the primary tumour's organ of origin and histopathology are the strongest determinants of its clinical behaviour, but in 3% of cases a patient presents with a metastatic tumour and no obvious primary. Here, as part of the ICGC/TCGA Pan-Cancer Analysis of Whole Genomes (PCAWG) Consortium, we train a deep learning classifier to predict cancer type based on patterns of somatic passenger mutations detected in whole genome sequencing (WGS) of 2606 tumours representing 24 common cancer types produced by the PCAWG Consortium. Our classifier achieves an accuracy of 91% on held-out tumor samples and 88% and 83% respectively on independent primary and metastatic samples, roughly double the accuracy of trained pathologists when presented with a metastatic tumour without knowledge of the primary. Surprisingly, adding information on driver mutations reduced accuracy. Our results have clinical applicability, underscore how patterns of somatic passenger mutations encode the state of the cell of origin, and can inform future strategies to detect the source of circulating tumour DNA.</p>}},
  author       = {{Jiao, Wei and Atwal, Gurnit and Polak, Paz and Karlic, Rosa and Cuppen, Edwin and Danyi, Alexandra and de Ridder, Jeroen and van Herpen, Carla and Lolkema, Martijn P and Steeghs, Neeltje and Getz, Gad and Morris, Quaid D and Stein, Lincoln D and Al-Shahrour, Fatima and Zhang, Junjun}},
  issn         = {{2041-1723}},
  keywords     = {{Computational Biology/methods; Deep Learning; Female; Genome, Human; Humans; Male; Mutation; Neoplasm Metastasis; Neoplasms/genetics; Reproducibility of Results; Whole Genome Sequencing}},
  language     = {{eng}},
  month        = {{02}},
  publisher    = {{Nature Publishing Group}},
  series       = {{Nature Communications}},
  title        = {{A deep learning system accurately classifies primary and metastatic cancers using passenger mutation patterns}},
  url          = {{http://dx.doi.org/10.1038/s41467-019-13825-8}},
  doi          = {{10.1038/s41467-019-13825-8}},
  volume       = {{11}},
  year         = {{2020}},
}