Combining genomic data from tumours found in different organs has revealed previously unknown cancer-related genes, and has led to a system for classifying tumours that could predict a patient's response to therapy.

In recent years, studies of cancer genomics have focused on mapping the genetic changes that contribute to the disease. Rameen Beroukhim of the Broad Institute in Cambridge, Massachusetts, and his colleagues analysed how the number of copies of genes varied across almost 5,000 tumours representing 11 kinds of cancer. The team found 140 genome regions in which copy-number variations were associated with cancer. Of those, 102 contained no known cancer-related gene, suggesting a suite of new cancer drivers.

In related work, Chris Sander and his colleagues at the Memorial Sloan-Kettering Cancer Center in New York used data from 12 cancer types to group tumours on the basis of genomic signatures rather than according to the organ in which the tumour originated. These signatures might one day be used to personalize cancer therapies.

Nature Genet. 45, 1127–1133; 1134–1140 (2013) Footnote 1