Collection |

TGCA pan-cancer analysis

Genomic alterations in diverse cell types at different sites in the body give rise to hundreds of different forms of cancer, and the ways in which these changes result in tumors with different biology, pathology and treatment strategies are beginning to be characterized. The TCGA Research Network has catalogued aberrations in the DNA, chromatin and RNA of the genomes of thousands of tumors relative to matched normal cellular genomes and has analyzed their epigenetic and protein consequences. Here the Pan-Cancer initiative examines the similarities and differences among the genomic and cellular alterations found in the first dozen tumor types to be profiled by TCGA. This first look across cancer types offers new tools in genomics and bioinformatics and the prospect of repurposing targeted therapies directed by the molecular pathology of the tumors in addition to their clinical classification.

Threads

Research articles

Current clinical practice is organized according to tissue or organ of origin of tumors. Now, The Cancer Genome Atlas (TCGA) Research Network has started to identify genomic and other molecular commonalities among a dozen different types of cancer. Emerging similarities and contrasts will form the basis for targeted therapies of the future and for repurposing existing therapies by molecular rather than histological similarities of the diseases.

Commentary | Open Access | | Nature Genetics

Chris Sander and colleagues have extracted significant functional events from 12 tumor types. Tumors can be classified as being driven largely by either mutation or copy number changes, and, within this division, subclasses of cross-tissue patterns of events are discerned that suggest sets of combinatorial therapies.

Analysis | Open Access | | Nature Genetics

Rameen Beroukhim and colleagues analyzed somatic structural alterations in 12 tumor types. Whole-genome doubling was found in over a third of all cancers, associated with TP53 mutation. Fifteen new significantly mutated candidate driver genes were found associated with recurrently amplified or deleted regions.

Analysis | Open Access | | Nature Genetics

Cancer genomic approaches have identified scores of genes responsible for the initiation and progression of cancer. But as the sample sizes increase, the list of putatively significant genes identified by current analytical methods continues to grow and is likely to include many false positives. This study shows that this situation stems largely from mutational heterogeneity and presents a novel methodology, MutSigCV, that overcomes the problem by incorporating mutational heterogeneity into the analysis. Application of MutSigCV to more than 3,000 tumour samples from 27 different tumour types shows that mutation frequencies vary more than 1,000-fold between extreme samples both between and within tumour types. And when applied to a data set on lung cancer, MutSigCV reduced the list of significantly mutated genes from 450 to a more manageable 11, most of them previously reported to be mutated in squamous cell lung cancer.

Letter | | Nature

As part of The Cancer Genome Atlas Pan-Cancer project, these authors present data analysis for point mutations and small indels from more than 3,000 tumours representing 12 tumour types. Among the findings are 127 significantly mutated genes from cellular processes with both established and emerging links to cancer, and an indication that the number of driver mutations required for oncogenesis is relatively small. Additional analyses also identify genes with significant impact on survival and a likely temporal order of mutational events during tumorigenesis.

Article | Open Access | | Nature

Viruses contribute to the pathogenesis of certain cancers. Using massively parallel sequencing data from The Cancer Genome Atlas to analyse viral expression in 19 tumour types, Tang et al. both confirm and reject previously described viral associations and present new information on viral integration and host interaction.

Article | Open Access | | Nature Communications

Analysis of data from The Cancer Genome Atlas generates a pan-cancer network of 143 recurrent miRNA-target relationships. The identified miRNAs were frequently regulated by genetic and epigenetic alterations in cancer. The work also reveals that some miRNAs might coordinately regulate cancer pathways, such as miR-29 regulation of TET1 and TDG mRNAs, encoding components from the active DNA demethylation pathway.

Resource | Open Access | | Nature Structural & Molecular Biology

This paper reports a strategy for combining somatic mutation profiles of human tumors with gene networks to stratify tumors into biologically and clinically relevant subtypes. The method is applied to ovarian, uterine and lung cancers.

Article | Open Access | | Nature Methods

Reuben Harris and colleagues report an analysis of gene expression and mutation data for multiple tumor types. They show that the DNA cytosine deaminase APOBEC3B is upregulated and that its preferred target sequence is frequently mutated in many types of cancer

Analysis | | Nature Genetics

Dmitry Gordenin, Gad Getz and colleagues report an analysis of mutation patterns in cancer genomes and find evidence of mutagenesis induced by APOBEC cytidine deaminase enzymes. They find an APOBEC mutagenesis pattern in bladder, cervical, breast, head and neck, and lung cancers, representing 68% of all mutations in some samples.

Analysis | | Nature Genetics

Larsson Omberg and colleagues write a Commentary describing the collaborative model used by the Pan-Cancer Working Group of The Cancer Genome Atlas. Pan-Cancer members used the Synapse software platform to share and evolve data, results and methods to perform integrative analyses of genome-wide molecular data for 12 cancer types.

Commentary | Open Access | | Nature Genetics

Additional articles

As part of The Cancer Genome Atlas Pan-Cancer project, these authors present data analysis for point mutations and small indels from more than 3,000 tumours representing 12 tumour types. Among the findings are 127 significantly mutated genes from cellular processes with both established and emerging links to cancer, and an indication that the number of driver mutations required for oncogenesis is relatively small. Additional analyses also identify genes with significant impact on survival and a likely temporal order of mutational events during tumorigenesis.

News & Views Forum | | Nature

The Cancer Genome Atlas, a large-scale genomics project to catalogue cancer-linked mutations, is starting to produce results. Glioblastoma, the most common brain cancer, was the first target for the project and the initial results, published AOP on 4 September, are now in print. Genes newly implicated in glioblastoma include tumour suppressors (NF1, RB1, ATM and APC) and several tyrosine kinase genes. Glioblastoma is extremely resistant to therapy, hence the potential importance of the development of a possible model system. Zheng et al. report that mice lacking the tumour suppressors p53 and Pten develop tumours resembling human glioblastomas, associated with increased Myc protein levels. As well as offering a potential system for testing therapeutics, this points to c-Myc as a possible drug target.

Article | | Nature

The Cancer Genome Atlas (TCGA) project reports here its analysis of messenger RNA and microRNA expression, promoter methylation, DNA copy number and exome sequences in 489 high-grade serous ovarian adenocarcinomas. The analyses help establish new tumour subtypes. Among other insights is the finding that while the gene encoding p53 tumour suppressor is mutated in almost all tumours, nine other loci including NF1, BRCA1, BRCA2, RB1 and CDK12 carry recurrent albeit low-prevalence mutations. Homologous recombination is defective in about half of the tumours studied, and Notch and FOXM1 signalling are involved in the pathophysiology.

Article | | Nature

The Cancer Genome Atlas consortium reports on their genome-wide characterization of somatic alterations in colorectal carcinoma. Combined analysis of exome sequence data, DNA copy number, promoter methylation, messenger RNA and microRNA expression, as well as low-coverage whole-genome sequencing reveal that 16% of these carcinomas have hypermutation. Of these, three-quarters have the expected high microsatellite instability, but the rest have somatic mismatch repair gene mutations. The data reveal a remarkably consistent pattern of genomic alteration, with 24 genes being significantly mutated. As well as identifying new biomarkers for aggressive colorectal carcinoma, the data imply an important role for MYC-directed transcriptional activation and repression.

Article | Open Access | | Nature

The Cancer Genome Atlas consortium has analysed 178 lung squamous cell carcinomas, a common type of lung cancer for which comprehensive genomic analyses have not previously been available. The researchers report that this tumour type is characterized by complex genomic alterations, with recurrent mutations in 18 genes, including TP53 in nearly all samples. They also report frequent mutations in squamous differentiation genes. Collectively, these analyses identify potential therapeutic targets worthy of further investigation.

Article | Open Access | | Nature

This Article from the Cancer Genome Atlas consortium describes a multifaceted analysis of primary breast cancers in 825 people. Exome sequencing, copy number variation, DNA methylation, messenger RNA arrays, microRNA sequencing and proteomic analyses were performed and integrated to shed light on breast-cancer heterogeneity. Just three genes — TP53, PIK3CA and GATA3 — are mutated at greater than 10% frequency across all breast cancers. Many subtype-associated and novel mutations were identified, as well as two breast-cancer subgroups with specific signalling-pathway signatures. The analyses also suggest that much of the clinically observable plasticity and heterogeneity occurs within, and not across, the major subtypes of breast cancer.

Article | Open Access | | Nature

This paper from The Cancer Genome Atlas Research Network presents an in-depth genome-wide analysis of endometrial (uterine) carcinomas from more than 350 patients. Based on a series of genomic features including newly identified hotspot mutations in the DNA polymerase gene POLE, and novel mutations in the ARID5B DNA-binding protein, the authors propose a reclassification of endometrial tumours into four distinct types. This might have clinical relevance for post-surgical adjuvant treatment of women with aggressive tumours.

Article | Open Access | | Nature

The Cancer Genome Atlas consortium reports an integrative analysis of more than 400 samples of clear cell renal carcinoma on the basis of genomic, DNA methylation, RNA and proteomic characterization. The data reveal frequent mutations in the PI(3)K/AKT pathway, suggesting that this pathway might be a potential therapeutic target, in addition to an array of epigenetic alterations that are linked to specific mutations in chromatin-associated proteins. One notable finding is the presence of a metabolic shift in aggressive cancers, correlating with tumour stage and severity.

Article | Open Access | | Nature