Identifying common genes is only the start. Some mutations seem to play starring roles in one cancer while being supporting characters in another. For example, Dr Ding discovered that the BRAF mutation, known to be present in about half of melanomas, was also found in other cancers, such as 7% of lung adenocarcinomas and 4% of colon and rectal carcinomas.

Even though there remains more basic science to be explored, the genetic approach points to a new way of matching patients to drugs. At the moment, many targeted cancer drugs are accompanied by a diagnostic test for the relevant mutation. The falling cost of sequencing means that a patient might now be tested for hundreds of mutations or even his full “exome” (the DNA that encodes proteins, less than 2% of the total). This means that someone who became resistant to one treatment could immediately be given another, based on the results of such tests. He might even be given a cocktail of treatments in the way that a cocktail of antiretroviral drugs is used to suppress HIV.

Cheaper sequencing is also changing the way that clinical trials are conducted. In November a group of academic researchers and companies, in collaboration with America’s National Cancer Institute, announced the Master Protocol trial. This will examine five potential treatments for squamous-cell carcinoma, a type of lung cancer. Rather than hunting separately for five groups of patients with the mutations relevant to each drug, the researchers will use a test from Foundation Medicine, a company that screens patients for abnormalities in 236 genes linked to cancer. They will then match volunteers with each of the five drugs. This approach cuts the administrative time and hassle of what would otherwise be a huge undertaking.