I know it sounds strange, but I feel very grateful to a database. It saved me from a lifelong fear of dropping dead of a heart attack.
The database is known as ExAC, and I had my first experience with it after I got my genome sequenced. For a few weeks, I brought it from one lab to another to ask scientists to help me make sense of it.
Their analysis brought up a doozy of a finding. I have a variant in a gene for heart muscles, called DSG2. Some studies have indicated that having a variant in just one of your two copies of DSG2 can cause a rare condition called arrhythmogenic right ventricular cardiomyopathy. It leads to sudden, lethal heart attacks; there is no cure.
I learned about my variant from Dr. Robert Green, a clinical geneticist at Harvard Medical School, and his colleague Matthew Lebo, the director of bioinformatics at the Laboratory for Molecular Medicine at Partners HealthCare. As they told me about its terrible reputation, a dark cloud of dread rolled over me as I learned more about DSG2.
But then—with the help of ExAC—Green and Lebo swiftly erased the cloud from the sky.
ExAC (short for the Exome Aggregation Consortium) is a compendium of genetic information that went online in October 2014. It contains every variation in every protein-coding gene in over 60,000 people. It’s about 10 times bigger than the previous database of its kind and contains about 10 million variants—so big that it is giving scientists a much more realistic vision of how we differ from each other genetically.
Green and Lebo visited the ExAC website to evaluate my DSG2 variant. They could instantly see that around 1 in 200 people of European descent carry it as well. That rate is far higher than the rate of arrhythmogenic right ventricular cardiomyopathy.
Green and Lebo diagnosed my variant as likely being benign.
ExAC has been steadily attracting a growing audience of scientists and doctors since its launch, mostly through word of mouth. For doctors like Green, it’s proving valuable as a way to find false positives. He has had the pleasure of telling some of his patients that a genetic diagnosis they got years ago is no longer a cause for concern.
“This resource alone is reshaping how we are doing this,” Green said.
But a set of new papers published Wednesday—the first in-depth analysis of the ExAC database—shows that false positives are just one of many things that it can illuminate. It is also helping scientists identify new genetic variants and understand secrets about the fundamental biology that we all share.
“This kind of tool is going to be extraordinarily important in all sorts of scientific discoveries,” said Green, who was not involved in any of the new studies.
False positives are an unfortunate legacy of the early days of genetic research. Until recently, sequencing DNA was so expensive that geneticists could only look for disease-causing variants in small groups of people.
In a typical study, they would examine the DNA of someone with a disorder and discover a suspicious mutation. If you suffer from cardiomyopathy, for example, a mutation in a heart-muscle gene may immediately send up a red flag. Then the geneticists would look at perhaps 100 healthy people. If they didn’t share the mutation, the scientists would publish a case report, pointing to the mutation as a possible cause.
But it could very well be that healthy people were walking around with that same variant. The geneticists missed them because their study was too small. Nevertheless, these weak reports have made it into many of the databases that geneticists use to diagnose patients.
“There’s a basic cleanup that needs to be done, because of the general crappiness of the databases that came before,” said Dr. Jay Shendure, a geneticist at the University of Washington who is not a member of the ExAC team.
Writing in the journal Nature, the team that built ExAC provided a 30,000-foot view of that general crappiness. On average, the researchers found, each ExAC subject has 54 variants classified as causing a genetic disorder—almost all of which are probably false positives.
My own false positive, it turns out, is part of a bigger story. In a separate study published Wednesday in Genetics in Medicine, the ExAC team collaborated with a group of British cardiologists to look at the genetic roots of cardiomyopathy. They compared the 60,000 people in ExAC to 7,085 cardiomyopathy patients who have had their protein-coding genes sequenced.
The scientists identified many variants like mine. Although they’ve been linked to cardiomyopathy, they are too common in healthy people to pose a genuine risk.
“We need to bring people up to speed on this,” said James Ware, a cardiologist at Royal Brompton Hospital and Imperial College London and a coauthor of the study. “Your cardiologist is going to need to understand it.”
For Ware, ExAC is useful for more than just finding false positives. It’s also helping him discover new disease-causing variants.
In their new paper, Ware and his colleagues describe how they found a small region of a heart muscle gene called MYH7 that’s a “hotspot” for cardiomyopathy. A fairly high fraction of people diagnosed with the disease has mutations in the region. But an unusually low fraction of the people in the ExAC database has variations there.
Based on that big difference, Ware and his colleagues estimate there’s a 98 percent chance that a mutation in the MYH7 hotspot will lead to cardiomyopathy—even if no one has seen that mutation before.
“We’ll be able to make a confident diagnosis more often,” said Ware.
ExAC’s importance extends beyond the clinic, and into labs where scientists are studying the basic biology of our genomes. In another study, the ExAC team searched their 60,000 subjects for broken genes—those with variations that are so devastating that they no longer make a protein at all.
The new study showed that broken genes are surprisingly common. Each ExAC volunteer had, on average, 85 genes in which one copy is broken. That same individual also carried 35 other genes in which both copies are broken.
Different people had different broken genes, the researchers found. But when the ExAC researchers catalogued all the genes in ExAC, they found that 3,230 genes were almost never broken.
“These are genes you can’t mess with,” said Daniel MacArthur, a principal investigator on ExAC and a geneticist at the Broad Institute.
Losing even a single copy of one of these vital genes may be lethal even before birth, MacArthur said. Breaking other genes may allow people to live to adulthood, but may cause such serious diseases that they have relatively few children who can pass their mutation on to future generations.
That’s just a hypothesis for now. The startling fact of the matter is that of these 3,230 genes you can’t mess with, 72 percent have no known link to any disease. In fact, scientists aren’t really sure about what their function normally is.
“It’s shocking that there are thousands of genes that are important, and we still don’t know what they do,” said Shendure, the University of Washington geneticist. But at least now, thanks to ExAC, they at least know just how important those mysterious genes are.