- Eat & Drink
- News & Features
- City Life
- The Hamptons
- Las Vegas
- Los Angeles
- New York
- Orange County
- Palm Beach
- San Diego
- San Francisco
- Silicon Valley
- Washington, D.C.
UCSF Knows When You'll Die
Scott Lucas | Photo: Janis Christie/Getty Images | March 28, 2013
...Or at least which chromosomal material might kill you. All this from a stockpile of spit.
Take 100 scientists, give them six years, $25 million, and the DNA-rich spit of 110,000 people, and what do you get? Answers. A ton of them. For the all-star lineup of researchers and doctors at UC San Francisco and Kaiser Permanente, that was the idea behind the joint Research Program on Genes, Environment, and Health, which launched in 2007 with the mission of collecting as much data about living, breathing humans as they could get their latex-gloved hands on. There would be so much information streaming in—an “informatics dream,” as Neil Risch, a UCSF genetic epidemiologist and the project’s co-leader, calls it—that the team’s computers would be literally overwhelmed. But the results of all the data-crunching would be stunning, and, as we’re beginning to see now, they could revolutionize not just how we treat disease, but also how we understand all of human suffering.
Take, for example, the findings on telomeres (sequences at the end of chromosomes that shorten with each replication). Telomeres are the body’s doomsday clock, linked to disease and now, thanks to the Kaiser-UCSF study, to mortality: That is, the shorter your telomeres, the more likely you are to die. UCSF’s Elizabeth Blackburn, who won the Nobel Prize in 2009 for her work in the field, oversaw the telomere measurement process; other researchers took those numbers and analyzed the hell out of them. It turns out that factors ranging from air and water pollution to crime rates and access to park space all affect the rate at which telomeres shorten and the body ages. According to Risch, even education level seems to affect their length. In other words, poor people don’t die younger just because of the cruel conditions of poverty—they die because they’ve been genetically compromised.
It’s revelations like these that are making the researchers understand just what a scientific gold mine they’ve uncovered. And they’ve done it with record speed. Whereas the first human genome took 13 years to decode, the UCSF-Kaiser collaboration has mapped the genomes of what amounts to the entire population of Berkeley in a mere 14 months. Patients who gave their consent received a saliva kit that looked like a large lens case. “They spit in the kit,” says Kaiser’s lead investigator, Cathy Schaefer, “and mailed it back to us.” Then scientists analyzed the DNA samples using technology developed by the Santa Clara– based biotech firm Affymetrix and tailored to Northern California’s diverse population: different arrays for patients of European, East Asian, African-American, and Latino ancestry.
Now researchers are comparing the 70 billion genotypes collected during the mapping process with the medical records of the patients from whom those genes came. Kaiser has been keeping electronic records since 1995 (long before, say, UCSF)—details of immunizations, blood tests, urine samples, mental health diagnoses, x-rays, MRIs, and on and on. The genome study includes 12,000 people with diabetes, 12,000 with depression, and 14,000 diagnosed with cancer—with records documenting family history, treatments, and health habits such as smoking, diet, and exercise. All of that information will be crunched and re-crunched.
What has the bio-geeks most excited is how the data (“de-identified” to protect patients’ privacy, Schaefer says) can potentially be used to treat and/or prevent all kinds of ailments and conditions. Take cholesterol, a key culprit behind heart disease and related conditions from stroke to dementia. Scientists have long known that lifestyle and environment affect cholesterol levels, but they didn’t know what role heredity plays. Now they’ve discovered links to 83 genes, against which they can analyze a million cholesterol tests (the average patient in the study has had 10). Before long, the team will be able to pinpoint which genes control reactions to different types of cholesterol-lowering drugs. Physicians will be “checking to see if their patients have this genetic variant or that one, and then prescribing this medication or that one,” Schaefer says.
The Kaiser-UCSF team expects to see these kinds of discoveries for decades to come, as the project opens its data to researchers around the globe. Meanwhile, scientists continue to collect and analyze samples from more patients—200,000 so far. Someday, they hope, almost everyone in Kaiser’s vast system will consent to being genetically mapped—and who knows what they’ll find then.
Originally published in the April 2013 issue of San Francisco.