Paul Allen's "brain atlas" unlocking mysteries of the ultimate computer

The type of brain map that used to grace high-school biology texts looked like a quilt: A pink chunk labeled "vision" bumped up against a blue blob that was the seat of language and a yellow swath representing motor perception.

Those crude representations were the result of centuries' worth of painstaking dissection, coupled with case studies of people suffering from brain damage and disease.

It took only three years for a Seattle lab founded by Microsoft mogul and philanthropist Paul Allen to revolutionize the landscape of neuroscience by creating a map of the brain that goes far beyond topography to pinpoint the workings of individual cells.

Allen, who donated $100 million to the lab, said he is so pleased with the results that he will consider similar, large-scale science projects in the future.

"This was a great opportunity to do something here that made a difference, and that we could do quickly," he said. "We'll certainly look for more opportunities like this."

Experts say the Allen Brain Atlas, which will be formally unveiled today, will boost understanding of brain circuits and chemistry — and what goes wrong in conditions ranging from schizophrenia and autism to Parkinson's disease and drug addiction.

The map is of the mouse brain, but that doesn't mean it's not applicable to humans. The two species share nearly 99 percent of their genes, and much brain research is conducted in mice.

Neurosurgeons at Swedish Medical Center in Seattle are already using the atlas to study the genetics of fatal brain cancers.

"This really just bolts us ahead in our ability to understand brain function and brain disorder," said neuroscientist Marc Tessier-Lavigne, who discovered the molecules that trigger connections between nerve cells, and serves as a science adviser for the Allen Institute for Brain Science. "It's cool, and it's important as well."

Allan Jones, the institute's chief scientific officer, likens the advance to leaping from a 19th-century map of Seattle that showed land masses and bodies of water, to zooming in on a single house with Google's interactive mapping system.

The project builds on the explosion of information from gene sequencing projects, like the Human Genome Initiative. Having lists of genes is valuable but doesn't necessarily reveal much about what those genes do in the body — or where they are active.

The atlas fills in those blanks for the brain, the body's most complex organ.

Using a robotic system to analyze 16,000 paper-thin mouse brain slices a week, the institute determined where in the brain each of the mouse's 21,000 genes is switched on, or expressed.

Almost every cell of the mouse body contains a full complement of all genes. What a particular type of cell looks like and how it functions is determined by which genes are switched on, so that a liver cell is different from a skin cell — or a brain cell.

The gene expression map provides unprecedented detail about the types of neurotransmitters and other chemicals produced in different regions of the brain and in different types of brain cells.

"The way brain cells are programmed to do their job is through the expression of specific genes," said Tessier-Lavigne, vice president of research drug discovery for the biotech firm Genentech.

Scientists were surprised to discover that 80 percent of all mouse genes are switched on somewhere in the brain, Jones said.

"We think there are probably more cell types within the brain than in all the other organs of the body combined," he said.

The free, online mouse atlas has been getting 250 hits a day from researchers working on a wide range of neuroscience problems.

"I really can't live without it," said Ben Barres, professor of neurobiology and developmental biology at Stanford University. He studies a little-understood class of brain cells called glia. The atlas allows him to quickly confirm which genes are switched on in various types of glial cells.

Mapping the genes himself would be far too time-consuming, he said.

Researchers studying Parkinson's disease can use the atlas to learn more about the genetics of the brain cells that are destroyed by the disease, leading to tremors. Multiple-sclerosis experts can explore why the crippling disease attacks some parts of the brain more than others.

As a programmer, Allen said he has long been fascinated by the ultimate computer: the human brain. When he was considering a brain project, he convened a meeting of top neuroscientists to find out what would be most valuable.

The resulting atlas is one of his biggest philanthropic endeavors. And though it lacks the pizzazz of some of Allen's higher-profile projects, like the search for extraterrestrial life or his X Prize-winning Space Ship One, it similarly blends the billionaire's interests in science and technology.

The atlas also generated staggering amounts of data, and much of the project focused on computer programs to make the information easily accessible.

The federal government would have been reluctant to fund a project on such a scale, said Dr. Thomas Insel, director of the National Institute of Mental Health.

"For a private donor to build this kind of a spectacular electronic atlas and just say to the world: 'Have at it' — I can't think of a precedent for that," he said.

Completing the atlas cost about $40 million, leaving the institute $60 million while it works to become self-sufficient. Jones said he hopes to bring in federal grants and forge partnerships with other scientists. The only outside money the institute has received so far is a $1.8 million grant from the U.S. Army to study the brain effects of sleep deprivation.

With its remaining Allen money, the institute plans to map the human neocortex — the wrinkled outer layer that is the seat of higher thought and emotion. Jones said they'll use cadaver brains and pea-sized bits of living tissue excised by surgeons during brain tumor removal or aneurysm repair.

Dr. Gregory Foltz, a neurosurgeon at Swedish, is partnering with the institute to help zero in on the genetic abnormalities glioblastomas — a type of brain cancer that usually kills within 18 months. Understanding why some genes are switched off in the tumors may lead to better treatments, he said.

Sandi Doughton: 206-464-2491 or


Allen Brain Atlas:

Allen/Nature Neuroscience Gateway:

Allen Institute for Brain