When the geneticist Gonçalo Abecasis stood up in front of a group of scientists in 2007 and proposed sequencing 1,000 genomes from people all over the world, he had no idea how he was going to pull it off. The Human Genome Project had published the first complete map of the human genome just four years earlier, and the technology remained exorbitantly expensive. “Imagine you’ve done ten of these,” Abecasis says today of the number of human genomes that had been sequenced at the time. “The first one cost $3 billion, and the other ones cost several million. And you say, ‘what if we set out to do a thousand?’”
The price wasn’t the only hurdle. “When we first started talking about doing these genome-wide comparisons between populations”—i.e., groups of people from different continents or with different ancestries—“there was a lot of tension,” remembers Abecasis, now at the University of Michigan. That’s because, frankly, “human population” sounds a lot like “race.”
Sequencing the human genome showed that humans are all much more alike than different; genome pioneer J. Craig Venter told the International Herald Tribune that the Human Genome Project proved once and for all that “the concept of race has no scientific basis.” And here was Abecasis, proposing that the next big international genomics project target precisely those differences and figure out what they mean. “There was some trepidation,” he says.
Now, eight years later, Abecasis’s idea—dubbed the 1000 Genomes Project—is coming to an end. They exceeded their goal, having sequenced over 2500 human genomes from 26 distinct groups of people around the world. (The collaboration’s last two papers are in Nature today.) Sequencing technology has improved by leaps and bounds since 1000 Genomes began, and the price for a whole genome sequence has fallen to a few thousand dollars. At least in the scientific community, trepidation about studying the genetic differences between groups of people has largely disappeared. In fact, many consider it the most important work they could possibly do. As the University of California, San Francisco sociologist Catherine Bliss puts it in her book Race Decoded, genomics has gone from being a “race-free” science to being a “race-positive” one.
To understand why, you have to understand what genetic variation is and what it does. Everyone’s genome is slightly different from everyone’s else’s, unless you’re an identical twin or, um, a clone. You inherited some of the differences from one or both of your parents; others might be random copying errors made when your genome was being pieced together. In their final 1000 Genomes paper, Abecasis and his colleagues estimate that a typical genome differs from the human reference genome—about 3 billion base pairs in all—in 4.1 million to 5 million places. It’s just a handful, really. But it's enough to make a difference.
Now, what happens if a relatively small group of people, with all their genetic quirks, gets separated from other populations? They’ll pass down those quirks to their children, and they’ll pass them along again, and so on. Eventually, those particular variations become more and more common in that isolated population, even though they might remain rare in the global population as a whole.
Those variations—also called alleles—might be neutral, like hair color. They might be beneficial in a particular environment, like those that allow Tibetans to adapt to high altitudes by lowering the amount of oxygen in their blood and helping them use it more efficiently. They might be detrimental, like the allele that makes Mexicans more susceptible to type 2 diabetes. Or they might be both. If you inherit faulty copies of the hemoglobin gene from both your parents, you’ll end up with sickle cell anemia. But if you only inherit one copy, that trait actually protects you from dying from malaria as a child. That double-edged sword might explain why the sickle cell allele became relatively common in Africa, where malaria is endemic.
The goal of 1000 Genomes was to catalog all of that variation, in hopes that, one day, scientists would be able to figure out what each change does—especially when it comes to disease. “If you said, 'I’m gonna take apart a car,' you’d get a list of parts. But if you want to understand, say, what makes some cars go really fast and other cars be really fuel efficient, you need to compare a few of them and see what the differences are,” Abecasis says. “Likewise, if you want to study disease, it’s not about [that fact that] we all have an enzyme that breaks down cholesterol. It’s learning that in some individuals, that enzyme is inactive, or in others it’s hyperactive, and trying to figure out what the consequences are.”
So in a way, race was a side exploration. The real point here might be genuine personalized medicine, using genetic quirks to find drugs, tests, and screenings that work on someone’s particular genomic makeup. The problem is that the vast majority of research into moving medicine in that direction study just one population: Europeans and European Americans, aka white people. “That’s a travesty, because huge parts of our world population are not deriving the same benefits of the fruits of the Human Genome Project,” said Esteban Burchard, a pulmonologist at UCSF, when I interviewed him about a study that mapped the staggering amount of genetic variation found in Mexico alone. Ignoring difference, in genomics as in society at large, doesn’t erase it. Instead, it guarantees that marginalized groups will be left out of what many scientists hope will be a medical revolution.
Abecasis and his colleagues ran across a concrete illustration of that disparity when they were analyzing 1000 Genomes’ last dataset. When they tallied up the number of genetic variants known to cause disease in each population they studied, they found significantly fewer of them in Africans than in people from Europe and Asia. That’s not because Africans don’t have as many disease-causing alleles. It’s because scientists haven’t looked for them. “A significant portion of biomedical research funding has been towards of populations of Eurasian descent,” so those are the genomes scientists understand best, says Adam Auton, a geneticist at the Albert Einstein College of Medicine and Abecasis’s co-author. “It’s kind of nice that by taking this global view of human genetic diversity, we can actually begin to see where some of these biases or gaps in our knowledge actually are,” he says.
That’s what Bliss, the sociologist, means when she says genomics has become a “race positive” science. “The idea is to help as many people as possible,” by targeting differences as opposed to glossing over them, she says. But she still thinks scientists like the ones involved in 1000 Genomes could do a better job clarifying what those differences actually mean. Agencies that fund biomedical research, like the National Institutes of Health in the US, want their money to help address health disparities between racial groups. A worthy goal, to be sure, but it often means scientists end up lumping genetically diverse groups of people together under one label that, biologically, doesn’t mean anything. 1000 Genomes sequenced people from 26 populations around the world, but, Bliss points out, the project also grouped them into the “super populations” of African, European, East Asian, South Asian, and Admixed American (Latin Americans with European and Native American ancestry, basically). “If you’re going to make comparisons, let’s not do continental comparisons,” Bliss says. Instead, compare each of the 26 groups on their own terms, without bothering with the charged idea of super populations. “Don’t wrap it back up in race,” she says.
1000 Genomes might not be perfect, but it has been an early step toward looking at human genetic variation in a way that recognizes difference but doesn't start and stop with race. Now, Auton says, it’s time for scientists to comb through the catalog they compiled and start figuring out what all that variation does—especially, which alleles are linked to disease. “It’s no longer about how we do this, it’s [about] what we do with it,” he says.
It’s easy to imagine a future full of routine whole genome sequences and personalized medicine, where 1000 Genomes’ focus on cataloging genetic differences between populations will seem blunt and antiquated, the product of an awkward transitional time in genomics. But by shifting the genomic conversation away from similarities and toward difference, Gonçalo Abecasis' longshot pitch for an ambitious genome study has lit a path to a different way of thinking about not just humans, but humanity.