Cancer is a disease of genes run amok, and scientists have found only a fraction of the bad actors. Tuesday, the government unveiled a $100 million project to speed discovery of culprits and cures, the first step toward a comprehensive map of cancer's genetic makeup.

It's an audacious project — the technology to even try it wasn't available just a few years ago.

And it comes at a crucial time: Half of U.S. men and one in three women will develop cancer in their lifetimes, and cases are poised to jump as the baby boomer population begins hitting 60 next year.

The Cancer Genome Atlas will "tackle the cancer problem like it's never been tackled before," said Dr. Francis Collins, genetics chief at the National Institutes of Health.

A cascade of critical genetic alterations — a domino effect — is required to cause any of the 200 diseases collectively called cancer. But despite 30 years of laborious work, scientists have found only a fraction of them. Those alterations differ by cancer type, and other genetic changes determine how aggressive each person's malignancy will be and even whether a particular treatment is likely to work.

"The challenge of cancer is its complexity," said leading cancer geneticist Dr. Ronald DePinho, of Harvard Medical School and the Dana Farber Cancer Institute. "There are a staggering number of genetic alterations that occur."

But the time is right for a massive, government-led push to unravel cancer's genetic underpinning because of recent successes mapping the human genome, a molecular blueprint of our species, said NIH Director Elias Zerhouni.

"Through this project I think what we will see is an acceleration of discovery," he said. "This is really the beginning of an era."

The project will bring researchers who now work independently together to hunt not just specific cancerous gene mutations, but chromosome rearrangements, faulty on/off switches and other abnormalities.

All the data from The Cancer Genome Atlas — abbreviated, in a bit of scientist humor, TCGA to reflect the four "letters" of DNA's code — will be made public for use by scientists anywhere in the world.

The first step is the three-year pilot project announced Tuesday, to focus on two or three cancer types, chosen within the next few months, to ensure the larger goal of a complete cancer gene map is technologically doable.

After all, it's a far bigger project than mapping humans' genetic makeup. "We're talking about basically thousands of Human Genome Projects," said Collins, who directed that program.

Even the pilot project will require collecting hundreds of tumor samples from hundreds of patients.

"We have the opportunity, because of advances in technology, to really look at the global nature of what is wrong with the cancer cell in a way that frankly we could not have dreamed of even a few years ago," Collins added.

The payoff could be huge. A handful of so-called targeted drugs — Herceptin, Gleevec, Iressa, Tarceva — are proving remarkably effective at battling certain cancers in patients with specific faulty genes.

That's only a small portion of patients: About 20 percent of breast cancer patients have the genetic aberration that Herceptin targets, and just 10 percent of lung cancer sufferers harbor the mutation that Iressa and Tarceva target.

But for those patients, the targeted therapies are proving effective — in Iressa's case doubling, to two years, some patients' survival, said Dana Farber lung cancer researcher Dr. Bruce Johnson.

Every new gene abnormality that's discovered is a potential target for another therapy.

"I personally have lost every member of my family to cancer. Every one is more difficult than the last, and I personally believe that this kind of project is going to change" treatment, said Dr. Anna Barker, a deputy director of the National Cancer Institute, which is running the new project with Collins' National Human Genome Research Institute.

While the project won't offer immediate new therapies, she acknowledged, "it's going to give a lot of hope to patients."