encoding the proteins that build our bodies and keep them functioning, play a crucial role in making us who we are—quite a massive haystack to go searching for needles in. But in theory, once the Human Genome Project collected and mapped out the genetic information of enough people, that haystack would suddenly become much easier to search. Now, all one would have to do is compare the genomes of a sampling of sick people—for any genetic disease, take your pick—with a control group, and whatever abnormality existed in the genome of the sick people would be impossible not to notice. Just like that, drug companies would have a gene to target—a specific genetic process to manipulate with medication.
With the Human Genome Project, new treatments and cures for any number of diseases seemed to be a few short years away. In 1995, the cancer researcher Harold Varmus, the director of the National Institutes of Health, organized a two-day workshop on schizophrenia at the National Academy of Sciences. Varmus, who with J. Michael Bishop had won a Nobel Prize for identifying the cellular origin of certain cancerous genes, had invited many of the usual suspects—E. Fuller Torrey, Irving Gottesman, Daniel Weinberger, Yale’s Patricia Goldman-Rakic—to present their latest research. Varmus was not impressed. At some point, Weinberger recalled Zach Hall, Varmus’s newly installed chief of NIH’s neurological disorders division, standing up and saying, “You people have been studying this disease for thirty years, and from where I sit, you have accomplished virtually nothing.”
Many of the researchers at the workshop were appalled. Some pushed back a little. Then Varmus himself weighed in, saying what he might have been planning to say all along. “You people don’t get it.” Everyone in the room could forget all about their enzyme research, their MRI studies, their CT scans, their PET scans. If you’re not studying genes, Varmus said, “you are going to be dinosaurs.”
* * *
—
COLLABORATING AROUND THE world, the scientists with the Human Genome Project thought the job would take fifteen years. They finished ahead of schedule, in 2003. Not only was the recipe book for human life now readable, beginning to end, for the very first time, but along the way scientists had identified new genetic markers spaced throughout the genome that could be used for research. Whereas DeLisi had previously been limited to about a few hundred different markers spread throughout the genome, the Human Genome Project opened the door to the discovery of literally millions more. With this wealth of new markers, researchers could now develop a tool for rapidly analyzing the genome, to home in on regions of DNA that seem to be associated with disease: the genome-wide association study, or GWAS.
The first step in a successful GWAS is to collect as many DNA samples as possible from people diagnosed with an illness of interest (for example, schizophrenia) and likewise to collect samples from a large number of apparently healthy people without that illness—the more samples, the better. With computer assistance, the GWAS method compares the information of these two groups, looking for any markers that are far more common among the people who are ill. Upon making this comparison, the theory goes, the genetic marker for any disease should be unmasked, almost instantly, for all to see.
In the first decade of the new millennium, there was a GWAS, and often more than one, under way for practically every disease suspected of having a genetic source: heart disease, diabetes, rheumatoid arthritis, Crohn’s disease, bipolar disorder, hypertension. In 2005, DeLisi was chairing a meeting of the International Society of Psychiatric Genetics in Boston when Edward Scolnick, a researcher from the Broad Institute of MIT and Harvard, announced that his institution planned to become the world’s clearinghouse for genetic data on schizophrenia, with an aim toward identifying schizophrenia genes with a GWAS. By 2008, virtually every researcher in the field, including DeLisi, participated in a new group called the Psychiatric GWAS Consortium (now the Psychiatric Genomics Consortium), which collected some 50,000 DNA samples from people with a range of psychiatric conditions, including DeLisi’s samples from the Galvin family. And in 2009, using information from that consortium, a study of 75,000 irregularities from more than 3,000 people with schizophrenia and bipolar disorder revealed “thousands of common alleles [possibly mutated genes] of very small effect.”
This psychiatric GWAS