The first people to set foot on Barbados, a wind-battered eastern spur of the Caribbean’s Lesser Antilles, came from the south, and relatively recently, no more than 1,700 years ago. Little remains of them: enough to know they were skilled farmers from the Orinoco Basin, in modern Venezuela. And like those of all humanity, their journey had started far earlier, when their ancestors, tens of thousands of years before, ventured out of Africa, across Asia, and into the Americas.
More people rolled in: The Lokano, clustered in scattered villages, hauling whelk and conch from the sea; and, in the 13th century, the Kalinagos, slipping on to the horizon in 50-foot-long dugout canoes. The Kalinagos reigned until the conquistadors. Hounded by European slavers, they fled windward to better defenses. By 1536, a Portuguese explorer could report Barbados as “uninhabited.”
It didn’t last. The English landed a century later and soon began importing slaves, ripped from their Ga, Igbo, and Ashanti communities in West Africa. By 1700, some 134,500 Africans lived in Barbados, in bondage; soon enough, 90 percent of Barbados’s population could claim African heritage, a percentage that holds true today.
A couple of decades ago, there was one more arrival: Kathleen C. Barnes, a graduate student and biological anthropologist from the University of Florida, who one day in 1991 walked into the emergency room of Barbados’s main hospital, the Queen Elizabeth. Throughout its human history, the island had had its share of plagues and troubles. Now Barnes was there to study a modern, quiet epidemic.
In a dedicated bay, child after child sat listless, worried mothers by their sides. The children were masked, inhaling medication for their wheezing, swollen airways. The machines hissed. Nearly one-fifth of Barbadians had asthma, far above the global average. Barnes wanted to find out why.
A native of a Virginia tobacco town known for housing the “Last Capitol of the Confederacy,” Barnes, who is white, grew up a witness to the civil-rights movement; in second grade, her school was forcibly desegregated. Trained initially as a nurse, she was troubled by the health disparities she saw in the United States. For example, African-Americans suffered from asthma far more than white populations did. There were many possible socioeconomic reasons. But Barnes thought it was mostly about pests.
Past research had tied some of the asthma rate in African-Americans to dust mites and cockroach feces, exposures that are more likely in poor communities. Barnes saw many similarities between African-Americans and the Afro-Caribbeans of Barbados, with one important caveat: Unlike residents of Baltimore, where she would go to work for decades as a professor at the Johns Hopkins University, the Barbadians had only just begun to live in homes conducive to household pests. A natural experiment had begun.
Barnes lived in Barbados for a year, running a lab across the street from the Queen Elizabeth, visiting homes to gauge their exposures. At the time, many Barbadians lived in chattel houses, movable wooden homes to which many residents had added bit by bit, enclosing them in concrete structures, with indoor plumbing. Dust mites loved the enclosed homes: Some of the levels Barnes measured were the highest ever recorded, she says. Surely that had to explain some of the asthma rate.
It probably did, as did other factors in a rapidly modernizing country: shifting diet and microbiome, rising obesity, wealth—the type of influences that are often lumped together as “environment.” But controlling for those, Barnes saw that a disparity still remained between people descended directly from Africa and those who came through Europe. Something more fundamental was at play, she realized. Something that would shape the next 20 years of her work.
“It seemed like the missing piece,” she said, “was understanding the genetic basis for these complex diseases.”
Genetics. Is there a more often misunderstood field? Put simply, it is the study of variation: how traits like height and risk of asthma are inherited, or not, through generations, human and otherwise. It can carry alluring, deterministic promises. And for as long as the discipline has existed, it has caused furious uproars.
The feuds started a century ago, after Gregor Mendel and his peas inaugurated the field. One camp, the Mendelians, argued that evolution came from discontinuous leaps, based on genes: Peas were either wrinkled or smooth, and inherited genes determined the difference. An opposing camp, the biometricians, argued that people are not like peas. We exhibit traits, like height, over a continuum. The first group was drawn to Mendel’s deterministic rules, a path that led some, assuming there could be a gene “for” a complex trait like intelligence, to eugenics. The other saw genetics as messy, interwoven, environmentally influenced.
They were all wrong, and a bit right.
The average person born today carries about 100 new mutations.
After several decades, a few scientists unified the tribes with a new theory, population genetics. Continuous variation could be explained by natural selection acting on differences among many genes. These differences, this evolutionary grist, come about through mistakes in repairing or copying DNA—insertions, deletions, and flipped letters in the code of life. We call these errors mutations. The average person born today carries about 100 new mutations among the six billion letter-pairs that make up our genetic code. We are all mutants.
Population genetics has always been ambitious, used to study all the genetic variation in and among species. Not all of this variation would be driven by Darwinian selection; through sheer randomness, new mutations that had no influence on reproductive success could win a place. Rapid population growth could throw a species out of balance, mutations accruing before selection weeded them out. Or a small group could migrate, losing access to a broader range of mutational variation, as the ancient Europeans once did; that’s why, even today, genetic diversity is lower in Europe than in Africa.
This work was largely theoretical. For many years, genetics became beholden to model organisms. Scientists inbred fruit flies, sucking the variation out of their DNA, creating little Mendelian machines. They lost touch with population genetics; it was as if, one researcher told me, they were working on airplanes without considering gravity and air.
The rise of cheap DNA sequencing over the past decade, and the failure of traditional genetic epidemiology—Barnes’s field—to detail the genetic basis of common disease, has led to a renaissance in population genetics. “We have way more data than we have models,” says Ryan D. Hernandez, an assistant professor at the University of California at San Francisco, and collaborator with Kathleen Barnes. The question now is: Can population and medical geneticists come to see the same world? Can they find the roots of complex diseases like asthma?
After graduating from the University of Florida, Barnes set out, in 1993, to burnish her genetic credentials, training as a fellow with David G. Marsh, an immunogeneticist at Johns Hopkins. Marsh was an ideal mentor, a founder of allergy genetics. He quickly became intrigued with Barnes’s stories of asthma on Barbados.
You have to go back, he told her.
This was before large-scale gene sequencing. Studies instead looked at families, using their patterns of inheritance, along with limited genetic markers, to infer important genes; such studies were traditionally used for diseases strongly tied to one gene, like cystic fibrosis. Marsh wanted to see how they’d work for a complex affliction like asthma, in which environmental influences—those dastardly dust mites, among other factors—are vital to the disease’s incidence, and inheritance patterns come out to frayed ends. But to do such work, they would first need to enroll extended clans of relatives—easily done in compact Barbados.
Asthma surprises you. It can lie dormant for years and then strike, spurred by some trigger. That happened to Harold Watson, an ER doctor on Barbados who has worked with Barnes for over a decade. He never suffered an attack until he was a university student. Then, one day, the airways to his lungs flexed, like a fist crunching a metal paint tube. He wheezed and seized up. It was terrifying.
“Your brain is totally clear,” Watson says. “You can breathe in, but you have difficulty breathing out. It’s like a clamp is holding you. You’re struggling with each breath to release that clamp.”
Cases like that spurred Barnes to search for what she’d thought would be the few genes that caused asthma. By the early 2000’s, she hoped that her work in Barbados and elsewhere, including studies of African-Americans in Baltimore, would allow clinicians to design more-effective interventions, diagnoses, and prognostics.
“We had this notion of one size fits all,” Barnes said. “That we would identify mutations or variants in genes. We’d hit on the key environmental factors. And that would be the answer for everybody.”
It didn’t turn out that way. After a decade, Barnes, and the rest of the genetics community, began to realize that family studies were an uncertain guide for common diseases, replete with false positives and few breakthroughs. Fortunately, a new solution seemed on the horizon: At great expense, scientists had sequenced the first human genome in 2001. They then developed chips, mostly based on European-descended populations, that could probe the most common variations in DNA for their statistical ties to a common, complex disease like asthma.
Why common gene mutations? Well, first it was all they could do, based on the technology of the time. But a simple idea also guided it: Because the diseases were common, it seemed intuitive that the DNA implicated would be widely shared. Barnes was one of the first to receive financing for such a test, called a genome-wide association study, in a population of African descent. Her asthma hunt was back on. Answers seemed imminent.
They weren’t. “The results were, frankly, disappointing,” Barnes says.
She had encountered a frequent problem in medical genetics, so common it’s a catchphrase: the “missing heritability.” Genome-wide association studies were producing robust, reproducible results, especially when done on a major scale, but the DNA regions they implicated often explained only a limited amount of a disease’s genetic roots. Something was missing.
In particular, Barnes had been watching a gene called ORMDL3, which was one of the few solid hits to come out of an association study on asthma in Britain. The gene helps gauge the inflammation of epithelial tissues lining the airways; the finding has pushed researchers to consider asthma as a disease of the epithelium. Barnes expected to see the same signal in her African-ancestry populations. But when she got the results and looked for ORMDL3, there was not a trace.
Let’s stop here to note: If you took any section of a person’s DNA and compared it with a stranger’s, no matter their ethnic background, odds are high they’d be identical. This is not platitude: These odds guide large-scale genome sequencing. They are fundamental. Humanity is deeply shared. It just happened that when it comes to asthma, for this one gene variant, people of European and African descent are distinct. At some point, after they diverged in ancient times, a mutation had taken hold. It wasn’t about race. It was about contingency. History.
“One thing we can’t do is use race as a proxy.”
“One thing we can’t do is use race as a proxy,” says Carlos D. Bustamante, a genetics professor at Stanford University and a Barnes collaborator. “It’s a very blunt tool. But we also can’t say there are no genetic differences across populations. Because it’s just not true.”
As Barnes mulled the ORMDL3 anomaly, genetics was debating where to go after association studies. There were many possible causes of the missing heritability: The tests could have missed a large number of common genes of very weak effect—so weak that they couldn’t pass statistical muster. Interactions among genes could cause it. Heritability estimates could be too high. Disease definitions could be broken; there could be many different asthmas, each with different genetic roots.
And there was one more possibility. Complex human disease could have roots in our own demographic history. In our population genetics. Our rapid expansion could be a source of our disease.
The first wave of large-scale genome-sequencing studies had begun to roll in. They showed humans riddled with rare gene variants: mutations shared by 1 percent or less of humanity. Sometime after the invention of agriculture, our population exploded—a trend that’s continued to this day. Mutations accrued. Think of our species as a giant family tree: More and more of the branches were surviving, sprouting their own mutations, and there was little time for these buds to spread. They would remain rare—private variation, as some call it now.
“It’s the single biggest result out of the last couple years,” says Joshua M. Akey, a genome scientist at the University of Washington, “the recognition of how dramatically the last 10,000 years have shaped patterns of human genetic variation.”
Many medical geneticists seized on rare variants as the next big answer. DNA-sequencing costs were falling through the floor. Surely looking at the genomes of thousands, to a high degree of detail, could solve the missing heritability? At the least, it could guide the next round of grants.
Indefatigable, Barnes began a new project: the first large-scale genome-sequencing study of African-ancestry populations, called Caapa (“Consortium on Asthma Among African-Ancestry Populations in the Americas”). Twenty-four groups contributed expertise or samples from 21 sites across the Americas—Barbados, Baltimore, Brazil. They were going to capture the common mutations that the European-based studies had missed. And their thousand genomes would be sequenced to a standard that few had previously reached in bulk, perhaps allowing them insight into the rare mutations that connected to asthma.
Hernandez, the UCSF researcher, had an even grander ambition for Barnes’s data. He had built a model of what could happen under rapid population growth, and it pointed toward signs that very rare, harmful gene variants could linger in the human population far longer than classical theory would project.
Hernandez needed a large number of full genome sequences to test his idea. He was particularly intrigued by the prospect of using sequences from individuals of mixed descent, like himself and the Caapa participants. Within each person’s genome lay a continuing experiment.
“Some individuals will have a stretch of their genome that’s European-American, but because of admixture, another chunk of the same chromosome, the same individual, will have African ancestry,” he says. “Or maybe they’ll have Native American ancestry.”
These DNA stretches experienced a different demographic history until 500 years ago, when the great, violent collision that we call the Americas began. By comparing these populations, Hernandez hoped to determine how much natural selection had acted on the genomes, along with gaining insight on the age of rare mutations. At the same time, working with Barnes, he might also expose how, say, a formerly helpful series of mutations found in African-descended populations might connect to asthma risk today. That was a favorite theory for Barnes, one she has held, in some form, for decades.
All they had to do was wait for the sequences to arrive.
The sequencing for Barnes’s new study was done by the time I visited her lab in late fall. It’s a ramshackle affair populated by blood-filled fridges and biochem benches, on Johns Hopkins’s hospital campus in East Baltimore. The DNA samples were sent to a private company, Illumina Inc., for sequencing, their data returned on a fleet of hard drives. The 1,005 high-quality genomes came to 130 terabytes. So much data is too big for the Internet’s pipes; it’s easiest to send through the mail.
Barnes had some early results to share, tucked among the Venn diagrams and colorful histograms that she slid onto her office table, though none yet about asthma risks. The study had found almost 50 million gene variants, and nearly half of those were new to science, unique to this study. A remarkable portrait of human variation had been missing, until now. Each site told its own story of mixture, reflecting its particular, often traumatic, history: the Puerto Rican subjects’ DNA was one-fourth of African descent, more than half European, and a tenth Native American; in Conde, Brazil, those splits were around 50-40-10. And in Barbados, more than 80 percent of the DNA had African ancestry, with the rest European; there were no Native American origins.
“Certainly we’ve developed a deeper appreciation for human variation, natural human variation,” Barnes said. A beautiful, diverse spectrum stretched out beneath her fingertips. “We all came from Africa,” she said. “And here we are today, in this continuum.”
Her team was sifting through these 50 million variants, searching for the most common to incorporate into a better gene chip. But deeper analysis of the full genomes has proved to be a challenge. Here’s the thing: Genomes are not sequenced in one long string of As, Ts, Cs, and Gs. They are reassembled out of fragments, and to fill in the missing information, early large-scale sequencing efforts, like the 1000 Genomes Project, relied on the probability that, at any point, one person’s stretch of DNA is identical to another’s. But the Caapa study sequenced its genomes 30 times over, and relying on that sort of genetic cross-stitching could potentially bias their results for those tantalizing, rare mutations. If you force people to look similar, you can’t see their differences.
“We don’t really know what to do, to some extent,” Hernandez now says. “Once you get looking at private variation—and now private variation ends up becoming, in a large sample, the vast majority of the variation—you get into tricky situations where it’s just difficult to figure out what’s right and what’s wrong.”
Over the past year, as they’ve begun to plunge into data sets like Caapa, population geneticists are starting to understand the implications of our many new, rare mutations—at least in theory. For example, as a species, we’re possibly no worse off after this rapid growth than if we had stayed at a flat number. Yes, we may have more unique harmful mutations, and natural selection has not had time to eliminate them. But our larger population makes it harder for these mutations to randomly spread, and more likely that older, harmful mutations from our pre-growth days will drop out.
“We’re way out of equilibrium,” says Andrew G. Clark, a population geneticist at Cornell University. Should our birth rate level off, “it will take thousands of years to come to balance. And that population will be much more diverse.”
Until then, life will be more of a genetic lottery for humanity. On average as a species, we’re doing fine, but there are winners and losers in the real world: some extended family trees that are relatively free of private, harmful mutations, and some that aren’t.
It also looks less likely that rare variants will be as important for complex disease as researchers hoped several years ago. Many, including Hernandez, assumed that the disease risk from a mutation would match a person’s evolutionary fitness—the likelihood that she or he would survive and have children. That’s far from a safe bet, subsequent work has shown.
Many diseases, like some forms of cancer and asthma, don’t arrive until adulthood, and may do nothing to harm a person’s likelihood of having children. In such cases, selection won’t weed those genes out, leaving them to become common or to vanish entirely. Other diseases, like severe autism, develop at a young age and could complicate the odds of having children; there, selection would act to keep those mutations down, making rare mutations likely to be more important.
Which diseases are driven by rare variants is “still an open question.”
The question then becomes, which diseases are likely to be driven by rare variants, and which ones aren’t? “And that’s still an open question,” Hernandez says.
A series of prominent experiments searching for rare disease-influencing mutations have had poor returns, several researchers mentioned to me confidentially, though many results have not been published. Stanford’s Bustamante puts the odds of rare mutations’ importance on a broad range, from 10 to 90 percent, depending on the disease. “It will run the spectrum, like everything in biology,” he says. And if most common diseases are linked to mutations of vanishingly small influence, there could be little geneticists can do, for now, to find these ties.
“These are the types of things,” Hernandez says, “that are just going to be really a huge pain to discover by genetic means.”
History repeats. Some researchers expect clear solutions; others expect more murk. Biometricians. Mendelians. There are many ways to stare at the verge of science. And if this story feels like a series of bait-and-switches, that’s because it is. The neat idea that initially drove Barnes’s interest in asthma genetics—that some sort of adaptive gene from Africa had gone haywire in modern contexts—has never appeared. Though it still could. There is still so much genetic and environmental material to be analyzed.
Now you can start to imagine what it’s like to be Kathleen Barnes. For her, this hasn’t been just a story. It is her career. She’s on a quest, but there’s been no dragon to slay. Just more road.
Paul Voosen is a senior reporter for The Chronicle.