They hope the work will enable them to replicate the successes of the few targeted cancer therapies such as Herceptin, useful against one type of breast cancer, and Gleevec, a drug that has revolutionized the treatment of chronic myeloid leukemia.

Scientists already know that cancer is a genetic disease, caused by mutations or other changes in the DNA of cells, but to date no one has done a systematic analysis of all the mutations in various tumours.

The Cancer Genome Atlas project launched by the National Cancer Institute and the National Human Genome Research Institute means to make a start on realising that goal.

"Cancer" describes a range of at least 200 different diseases and it is the second leading killer in most industrialized countries, after heart disease.

Modern treatments mean that now about 60 percent of patients are alive five years after the initial diagnosis.

Lung cancer or breast cancer are blanket terms for tumours that arise because of differing genetic mistakes.

Dr. Andrew von Eschenbach, director of the National Cancer Institute and the acting commissioner of the U.S. Food and Drug Administration, says the cancer genome atlas project is about making the necessary progress so that cancer is a chronic and manageable condition that no longer causes the suffering and death that is seen.

The researchers are as yet unsure which cancers they will tackle first but say it will be a small number.

They will need hundreds of samples from hundreds of patients with the selected tumour types and then use the human genome map made at NHGRI to try and find all the changes that mark cancer.

Apparently several cancer-causing mutations, called oncogenes, are known already.

Some of the best known are the BRCA1 and BRCA2 breast cancer genes, also implicated in some cases of ovarian cancer, the p53 gene involved in many different tumours, and the EGFR gene targeted by AstraZeneca's Iressa and Genentech and OSI Pharmaceutical's Tarceva, which have remarkable effects against a small percentage of patients with lung cancer.

The researchers suspect there must be many more and Dr. Ron DePinho of Harvard Medical School and the Dana Farber Cancer Institute in Boston believes they are dealing with the tip of the iceberg, as there are a staggering number of genetic alterations that occur.

DePinho says the project will "liberate the cancer community" from having to track down each gene one by one, and allow it to concentrate on what different genes do and which ones make the best targets for drugs.

Anna Barker, deputy head of the NCI, says the project is a turning point in biomedical research and for medicine.

Gene testing will allow doctors to find out immediately whether drugs already used against one form of cancer may be useful against others.

National Institutes of Health Director Dr. Elias Zerhouni, says being able to identify the precise genetic mutations in a tumour might make it easier to use intelligently designed combinations of drugs.