In the public imagination, cancer has a simple meaning. Whether it affects the breast, brain or bowel, this feared diagnosis is thought of as a single disease that always proceeds in much the same way. It means a body riddled with tumours, disfiguring surgery, gruelling chemotherapy and radiotherapy, hair loss and death.

In some ways, this is not too wide of the mark. Cancers are all caused by genetic mutations that promote unchecked cell division, and they kill when this rogue tissue invades vital organs. They are also all treated with the same three approaches, which attempt to cut, poison or burn out tumours. Each is a blunt instrument that causes side-effects by damaging healthy tissue.

This view, though, is out of date. Doctors have long considered cancer to be a group of diseases, not a single condition like measles or flu.

The full extent of its complexity should now be unravelled by an international study that was launched this week. It promises to transform scientific understanding of cancer, which could lead in turn to kinder and better new treatments.

The realisation that lung and liver tumours are not the same has already improved oncology. There are more than 200 types of human tissue: the cancers that affect each are subtly different in origin, progression and prognosis, and respond differently to the available treatments. A drug called cisplatin, for example, is quite effective against testicular cancer, but not against brain tumours.

This broader picture, however, is still too simple. It is now clear that even cancers of the same tissue can differ greatly at a genetic level, with important implications for therapy.

Some of these are already well known. About a quarter of breast cancers have a mutation in the HER-2 gene. These tumours are more aggressive, but they are also susceptible to treatment with the drug Herceptin. Breast cancer, indeed, can now be broken down into four or five genetic subtypes, each of which is handled differently in the clinic.

The same will be true of other cancers, which is where the new project comes in. The International Cancer Genome Consortium (ICGC) aims to map every genetic mutation that commonly occurs in 50 cancer subtypes. Some of these are unimportant, but others will be fundamental to these tumours' growth and spread. Many will also be present in cancers that affect very different tissues. A gene called BRAF, for example, is already known to be mutated in most of the deadliest skin cancers, and in many bowel cancers as well.

This knowledge will have two important implications. First, it will allow scientists to design new drugs that target only cells with particular damaging mutations, leaving healthy tissue untouched. Next-generation chemotherapy should thus be more effective, but with fewer side-effects.

Just as importantly, it will refine diagnosis well beyond the 200 or so kinds of cancer that are known today. Doctors will be able to test the DNA of their patients' tumours, and categorise them not as stomach or bowel cancers, but as cancers that are driven by particular mutations, such as BRAF. Therapy can then be tailored accordingly.

The ICGC is one of the most ambitious and expensive pieces of medical research that has yet been attempted. It will require more DNA to be sequenced than in the entire history of genetics, and it is anticipated to cost $1billion (£500million).

For cancers, though, it is worth thinking big. The understanding that these diseases are genetic in origin, and different, has for the first time given humanity powerful tools with which to beat them. That will not only prevent some of the 7.5million deaths that tumours cause worldwide each year; it should eventually change the popular perception of these conditions, lifting a little of the dread that the “C-word” always inspires.

Mark Henderson is the Science Editor of The Times

Explore Health

• Expert Advice

• Health Features

• Mental Health

• Alternative Medicine

• Child Health