CHERYL WHITTLE TRIED HER BEST to fall asleep, but her mind kept racing. Tomorrow was going to be the culmination of three years of research and, possibly, a day that would change her life forever.

Around four a.m. she popped two Benadryl and managed to drift off. But in just a few hours she had to be up and ready to go.

Cheryl and her husband, Dickie, are retired, and live in eastern Virginia, way out on the end of the Northern Neck peninsula, which juts like an arthritic finger into Chesapeake Bay. It’s a beautiful and isolated spot, where most people tack up “No Trespassing” signs and stay close to home. The Whittles enjoy their life in the country, but Cheryl was eager that day to make the long drive to meet Effie Jane.

She showered, threw on a T-shirt, jeans, and sneakers, dotted make-up on her cheeks, and scrunched a dollop of mousse into her thinning brown hair.

There’s nothing showy about Cheryl, not even on a day like this. She’s short and shy, with nine grandchildren and no pretensions. She grabbed a shoulder bag, heavy with the day’s supplies, and kissed Dickie on her way out the door.

Her anxiety mounted as she drove her yellow pick-up past sleepy cornfields, old plantations, and cemeteries, up the peninsula and into mainland Virginia. Then she pulled into the tiny parking lot of Panera Bread in Richmond. She didn’t have to wait long before Effie Jane Erhardt found her—that yellow truck was hard to miss. Effie Jane pulled open the truck’s passenger door and announced

“I’m here!”

Cheryl and Effie Jane met on Ancestry.com, a popular website for people trying to fill in their family trees. After several email and phone encounters, each woman felt a kinship that neither had experienced before. Both were born in 1951, and grew up about 20 miles from each other in the Richmond area. They both speak with soft Southern drawls, had traumatic childhoods, are devout Christians, and, as children, felt like outsiders in their own families.

Cheryl quickly got down to business, retrieving a small cardboard box from her bag in the back seat. She opened the top, plucked out a fat plastic tube, and handed it to her friend. Effie Jane held the tube under her mouth and spit—and spit, and spit, and spit. She had never realized how much saliva froths and fizzes. She passed the tube back to Cheryl, who snapped on a plastic cap, gently mixed the tube’s contents, and dropped it in a clear plastic bag with an orange BIOHAZARD label. Then the two women went into Panera for lunch.

SINCE 2000, when a company called Family Tree DNA sold the first commercially available home testing kit, an estimated one million people have dabbled in genetic genealogy—also known as recreational genetics, extreme genealogy, and even anthrogenealogy.

Traditionally, amateur genealogical research was regarded as a niche hobby for older white men, but today it attracts people of all ages, races, and walks of life.

The rapid transformation is due to two technological revolutions. Twenty years ago, doing genealogy meant hitting the pavement: traveling to local historical societies, courthouses, libraries, and cemeteries to paw through dusty books and records. Then came the internet, which made the most useful references—census and voter lists, birth certificates, military records, even the archives of local newspapers—accessible from home. Not only that, but genealogists started connecting with each other online, sharing their research and overlapping trees, creating a vast online database that anyone could tap into and, more importantly, add to.

In 1997, a company called Infobases, which sold compact disks of The Church of Jesus Christ of Latter-day Saints publications, bought Ancestry Magazine and its website, Ancestry.com, turning the latter into a subscription genealogy service. By 2009, when Ancestry.com went public, it had a near-monopoly on the booming industry. The world of ancestry research has become a perfect example of a highly scalable business based largely on freely provided, user-generated content. Today Ancestry.com has a few competitors, like MyHeritage.com and Brightsolid, but it remains dominant, with almost three million paying subscribers, 12 billion records, and 50 million family trees. Revenues from the company’s ten popular websites and the Family Tree Maker software totaled $400 million in 2011. In late 2012, a European private equity firm bought the company for $1.6 billion.

The second transformation came from rapid advances in genetic testing. After Family Tree DNA launched its test, other companies followed: 11 by 2004, and almost 40 by the end of that decade. Today you find celebrities like Meryl Streep and Yo-Yo Ma tracing their lineage on primetime television shows. As the price of commercial genetic tests has plummeted—many now cost just $99—families like the Whittles have been able to join in.

Three companies—23andMe, Family Tree DNA, and Ancestry.com—have emerged as major players, and each is intent on growing its most valuable asset: a proprietary database of customers’ genetic data. 23andMe has information from more than 400,000 people and counting, and Family Tree DNA has over 650,000 different genetic records.

The bigger these databases become, the more useful they are for filling in genealogists’ ever-expanding family trees. But the growth of the databases also raises serious privacy concerns—not only for people who buy the tests, but for close or even distant family members who share some of their DNA.

Searching your genetic ancestry can certainly be fun: You can trace the migration patterns of 10,000-year-old ancestors, or discover whether a distant relative ruled a continent or rode on the Mayflower. But the technology can just as easily unearth more private acts—infidelities, sperm donations, adoptions—of more recent generations, including previously unknown behaviors of your grandparents, parents, and even spouses. Family secrets have never been so vulnerable.

If you find a relative on a genetic genealogy database—say, a second or third cousin—then with the help of Google, social media, digital obituaries, and other publicly available resources, it’s usually possible to find closer kin. Adoptees have used their newfound genetic knowledge to browse photo albums and look for potential biological relatives on Facebook. Children of sperm donors have found siblings they never knew they had. Couples who used artificial insemination to conceive have discovered that another man’s sperm had been used.

And then there are people like Cheryl, who learn to their surprise, late in life, that they aren’t the person they thought they were.

Over sandwiches at Panera, Cheryl and Effie Jane exchanged photos and told childhood stories. Taking advantage of the restaurant’s Wi-Fi, Cheryl took out her laptop and logged into the 23andMe website, patiently explaining how the process worked. Cheryl was a veteran. Genetic testing had already shaken up her world, raising startling new questions about where she came from. She was here because she believed that Effie Jane was her sister. She was praying for it. If she was right, the journey she’d been on for the last three years would reach its end. Her mind could rest.

The women left the restaurant together, drove to a nearby post office, and sent the sealed package to a lab in Los Angeles. There, technicians would screen Effie Jane’s DNA for about one million genetic markers. Four to six weeks later, 23andMe would send Cheryl an email saying the results were ready.

CHERYL’S QUEST BEGAN one afternoon in late 2008, when she and Dickie were sitting in their living room watching Oprah. The episode included a segment about a Silicon Valley start-up called 23andMe that was selling genetic tests directly to consumers. One of the company’s founders, Anne Wojcicki, was nine months pregnant. She told Oprah how her DNA test results and those of her husband—Google co-founder Sergey Brin—offered clues about their unborn child.

Cheryl, then a registered nurse, was intrigued. She had married Dickie when she was 14 and he was 20. The pregnancy that spurred their young union resulted in a stillborn girl, born with too much fluid in her brain.

After hearing Wojcicki’s story, Cheryl thought that this DNA test might provide a genetic explanation for their daughter’s death. What’s more, she had been interested in genealogy for years, and had done a lot of work on the line of her father, Josiah “Joe” Wilmoth. Perhaps DNA testing would expand that research.

Cheryl remembered some of the basics about how genes work from nursing school, and learned more after browsing 23andMe’s website. For instance, most of us have 23 pairs of chromosomes, or long segments of DNA. Both sexes have 22 pairs of so-called “autosomal chromosomes”—each pair includes one copy inherited from each parent. But the 23rd pair, the “sex chromosomes,” is different. Men inherit a Y chromosome from their father and an X chromosome from their mother. Women, in contrast, receive an X chromosome from their father and a second X from their mother.

To investigate her father’s ancestry, Cheryl decided to look at the DNA of a male in her father’s line. Joe Wilmoth had died in 1989. But his son Milton Wilmoth, Cheryl’s older half-brother, was alive and well. So she called Milton and asked: If she paid for it, would he consider taking a genetic test?

Milton didn’t own a computer and had no interest in his DNA, but he readily agreed to help his sister. So, in early 2009, Cheryl bought three kits from 23andMe—one for her, one for her husband Dickie, and one for Milton—at $495 a pop. Two months after that, she was back at her computer poring through the results.

From what she could tell, nothing in her genes or Dickie’s gave any clues about why their first baby had died. But not long after she took the test, 23andMe launched Relative Finder, a service that allowed customers to find relatives in the company’s database based on shared segments from all 23 chromosomes. Oddly, Milton’s name did not appear in Cheryl’s list of DNA relatives. She tried signing into 23andMe using Milton’s account instead, and saw that her name did not appear in his list of DNA relatives, either.

After a couple of months, Cheryl reached out to CeCe Moore, an expert in genetic genealogy who runs a popular blog on the subject. “I wrote to her and said, ‘Can you tell me what I’m doing wrong? I can’t get this thing to work.’” Cheryl recalls.

CeCe revealed the truth that Cheryl suspected, but had been scared to confront: Milton was not biologically related to her. For Cheryl, there was only one explanation: She was a Wilmoth by name, but not by blood.

The notion that Joe wasn’t her biological father didn’t sit well with Cheryl. It meant, after all, that her mother had lied to her—and maybe to Joe and everybody else—for decades. But Cheryl didn’t feel anger toward her mother. In fact, the more she thought about it, the more she felt comforted. Many things about her childhood suddenly made sense. When she was growing up, Joe never gave her much affection or attention. And when he drank, which was often, he could turn mean. “I never felt a part of him,” she says. “I grew up believing in my heart that I did not belong.”

The news was a painful shock, however, to Cheryl’s younger sister, Sandi Satterfield, who was crushed at the thought that they weren’t full siblings. Joe had always doted on Sandi, and Sandi had adored him, despite his flaws. Yet she worried about what it might mean for her own roots. Could it possibly be that she, too, wasn’t his?

After prodding from Cheryl, Sandi agreed to take a 23andMe test to confirm that Joe was her father. When her results came back, in June 2010, they showed that she was at extremely high risk for colorectal cancer. Just seven months later she was diagnosed with stage IV of the disease. “If we had known this earlier in her life, she may have been able to take the appropriate actions to prevent this horrible disease,” Cheryl says.

But Sandi’s DNA results also resolved her worries about her own lineage. Cheryl and Sandi didn’t share half of her DNA, as full siblings do. They shared around 22 percent, making them half-siblings through their mother. Sandi shared 25 percent of her genome with Milton, which meant they were half-siblings through their father, Joe.

Sandi’s first reaction was acceptance. Four days after the results came back, she sent Cheryl an email.

But as the information sunk in, Sandi became distressed about the implications of the test.

The bottom line is that it really doesn’t matter. I love you and they loved you…in their own way I guess. Daddy was OUR daddy, nothing can change that.

She was hurt when, a few weeks after the results came through, Cheryl sent an email to Sandi and 15 other friends and family with the subject line, “Who is my biological daddy?”

In the email, she laid out the whole story of Milton and Sandi’s tests, revealing that Joe wasn’t her biological father and reflecting a bit on her feelings:

My thoughts…. Daddy Joe knew in his spirit I wasn’t his. After all, animals know their child from someone else’s, it is part of nature. AND male animals will usually destroy the offspring of other males. And momma never cared for me like she cared for Sandi, why? Because she had a problem with my father, whoever he was. Was she raped, was she involved with someone and he dumped her, or was she just ashamed for some reason. We do not know, as no one is here that we can ask.

Two days later, Sandi responded to the email and similar things Cheryl had posted on Facebook. She was upset that Cheryl referred to Joe as Daddy Joe, as opposed to just Daddy. “I know you mean no harm and [are] only trying to distinguish between sperm donor and daddy, but it really bothers me for him,” Sandi wrote. “I just feel so bad for him wondering if he knew and now I feel as though he was played a fool.”

“I should never have taken that test,” she added. “I feel so terribly guilty.”

BLAINE BETTINGER IS A WELL-KNOWN FIGURE in the genetic genealogy world. The 37-year-old is an intellectual property attorney by day. “But I joke that it’s just a way that I make money to pay for more genealogy tests,” he says. He has been researching his family history since he was a kid and studied molecular biology and genetics in graduate school, so he was perfectly poised for the genetic genealogy revolution. Bettinger bought his first genetic test in 2003. A few years later he launched a blog—the Genetic Genealogist—with the aim of explaining the science behind the tests in simple language. It now receives around a thousand visitors a day, he says.

Genetic genealogy can be extremely complicated, but most cases require only a basic understanding of our 23 pairs of chromosomes. Most chromosomes, you could say, are promiscuous. During the formation of egg and sperm, each chromosome inherited from the mother physically crosses with its counterpart from the father, and as the pieces mingle they freely exchange segments of DNA. This recombination gives our species great genetic diversity, and it’s the primary reason non-twin siblings are never genetically identical.

But the Y chromosome is chaste. The vast majority of its 50 million DNA letters do not swap with other chromosomes, passing almost identically from father to son, son to grandson, and so on. That means that when a genetic change spontaneously occurs in a Y chromosome, it can be passed down to male descendants forever, serving as a reliable marker of their paternal lineage.

This was famously demonstrated in 1997, when researchers published a study of Jewish priests in the journal Nature. According to Jewish belief, the high priesthood began 3,300 years ago with Aaron, Moses’s older brother, and has been passed from father to son ever since. Today, many Jews have the surname Cohen or Kohen, meaning “priest” in Hebrew. The researchers scraped a few skin cells from the inside the cheeks of almost two hundred Jewish men from Israel, North America, and England, and compared the mens’ Y-chromosomes. Close to 70 had been told at some point that they were direct descendants of the high priests. And these men, it turns out, had a distinctive Y-chromosome profile. “The simplest, most straightforward explanation is that these men have the Y chromosome of Aaron,” the lead researcher told the New York Times.

The following year, a similar genetic study made headlines when it bolstered the controversial theory that Thomas Jefferson had fathered a child with his slave, Sally Hemings. The researchers looked at the genes of male descendants of Jefferson’s paternal grandfather and found that they carried a combination of 19 genetic markers that is quite rare, showing up in just a tenth of a percent of all men. But the researchers found exactly the same set of markers in a descendant of one of Hemings’s sons, Eston, meaning that Eston’s father was either Jefferson or one of Jefferson’s close male relatives.

For genealogy buffs, these studies had thrilling implications. Since the Middle Ages, Western cultures have passed surnames from father to son. In theory, then, men who have the same surname should share markers on their Y chromosomes. This wouldn’t be true for everyone, of course: multiple families may have taken up the same surname even if they weren’t related, and adopted children often take the last name of their adoptive fathers. But it’s true for enough people to be useful for tracing family trees.

When Family Tree DNA launched its genetic genealogy test, which screened for 12 markers on the Y chromosome, genealogists could find members of their paternal line not with treks to libraries or cemeteries, but by uploading their DNA results to the company’s database.

That was 2000. By the end of 2001, the company’s customers had organized research projects for about a hundred surnames. After 23andMe launched, in 2007, it added thousands of markers associated with health risks, such as those that Cheryl heard about on Oprah. There are also companies that specialize in determining ancestry for African Americans, Native Americans, and other specific ethnicities.

For people like Bettinger, DNA testing has made genealogical research richer and more fulfilling. “Once I got the DNA test back, I was able to look at my family tree in a whole new light,” he says.

Bettinger is Caucasian and had assumed that his ethnicity was one hundred percent European. But tests revealed that he carried Native American markers. “This was a complete shock,” he says. He had known from his previous research some of his ancestors had lived in Honduras in the mid-1800s, but had assumed that they were all English missionaries. After getting his results back, he realized that some of them were native Hondurans, with ancestry from both Honduras and the Cayman Islands.

“Genealogy is not only about names, dates, and places, but about filling out the story of each ancestor as well—what their lives were like, what their motivations might have been like, the trials and tribulations and joys that they experienced in their lifetime,” Bettinger says.

“Every decision, no matter how small, by each one of these individual ancestors ultimately led to my existence—and, undoubtedly, to the person I am today.”

For some people who do genetic genealogy, though, the information they unearth is more difficult to accept. “You would not believe the things we can find out,” particularly when genetic information is combined with searches from the internet and social media, says CeCe Moore, the blogger who helped Cheryl. “If you’re a privacy advocate, it is worrying.”

Over the past three years, Moore says she has answered emails from more than ten thousand people interested in using genetic genealogy, and has intensely worked on searches for about a hundred people. Many of these people are adoptees or, like Cheryl, have discovered that they have mystery fathers. “We used to only have an adoptee get a close match every six months,” says Moore. “Now it’s happening every single week.”

“I believe that knowledge is power, and I think we can gain much more than we will lose from this movement,” says Moore. (She is also an unpaid liaison between several genetic testing companies and the genetic genealogy community, and a paid consultant for the popular US television show “Finding Your Roots with Henry Louis Gates, Jr.”) At the same time, though, there’s no denying that some of this newfound knowledge will be painful. “A lot of times people find out things that really shake their identity,” she says.

AFTER LEARNING THAT JOE WILMOTH WASN’T her biological father, Cheryl began unpacking what she knew about her mother, Vivian.

Vivian Tipton was strikingly beautiful, even in her older years, and had an infectious cackle of a laugh. She grew up in Petersburg, Virginia, a small town about 25 miles south of Richmond. In July of 1941, when she was 16, she married an 18-year-old soldier named Richard Thompson. Just five months later he left to fight in Europe.

Richard returned after the war, and he and Vivian moved into a house across from her parents. By the end of 1949, the couple had two girls, Toni and JoAnn, and Vivian was pregnant with a third. Vivian always said Richard was the love of her life, but their marriage was cut short on December 21, 1949, when Richard was killed in a dump-truck accident. Vivian was devastated, staying in bed all day and refusing to celebrate Christmas. After having her third daughter, Jayne, in February, Vivian and the girls moved in with her parents. Not long after, she moved out, leaving her children to be raised by their grandparents.

The next few years of Vivian’s life are not entirely clear, but sometime in 1950 or 1951 she met Joe Wilmoth. They married in the summer of 1951 and moved to Chester, about 10 miles from Petersburg. They had a rocky relationship, to say the least. Joe was physically abusive at times, and the couple seemed to never stop arguing. Cheryl was born less than seven months after their marriage, on Christmas Eve of 1951, and grew up believing that she had been a premature baby. After the DNA test, though, Cheryl wondered if even that were true. The test raised so many unsettling thoughts, the kind that kept her up at night. How many of the other stories of her early life, she wondered, were fiction?

From ages one to four, Cheryl lived with a couple who had grown up with Joe. Cheryl doesn’t know exactly why. It could be because Joe and Vivian weren’t getting along—or, perhaps, because Joe didn’t want to raise another man’s child.

The next decade of Cheryl’s life was unstable and traumatic. She lived in more than a dozen different homes in Florida and Virginia, and frequently witnessed violent outbursts from Joe. Looking back, Cheryl suspects that some of Joe’s behavior could be explained by post-traumatic stress disorder — he had seen combat in the Philippines during the war. As a child, though, no explanation would have helped. She was only terrified.

Through all of this turmoil, Cheryl tried to protect and care for her little sister, and the girls forged a powerful emotional bond. Still, they were different in more ways than they were alike. Sandi was tall and thin; Cheryl, short and plump. Sandi was happy-go-lucky from a very young age, and by the time she was a young woman, liked to drink, smoke, and party hard. Cheryl was shy, anxious, fearful, and prone to crying.

Perhaps their most striking difference, though, is in their relationship with Joe.

“To me, he was everything,” Sandi told me. “She was afraid of him.”

The day after Christmas, 1963, the family moved into a new home. The house was right up the road from Dickie’s family, and soon Cheryl and Dickie were sweethearts. She got pregnant in early 1966, soon after her 14th birthday, and they were married in May, just before Dickie turned 21.

Cheryl’s early marriage shows the extent of Joe and Vivian’s parental neglect, Dickie says. “It was pretty obvious,” he says. “I mean, you don’t let your 14-year-old girl go out with somebody as old as I was.”

In 1980, when Vivian was just 55, she was diagnosed with lung cancer.

Cheryl came over most days to make food and help clean the house. One day, she and Vivian opened up an old cedar chest of Vivian’s personal mementos. The chest contained a pink card issued by the hospital on the day Cheryl was born. It noted, in handwritten script, that she weighed almost seven pounds—much heavier than a baby who was two months premature could possibly be.

Vivian was sick for three years, and Cheryl’s relationship with Joe disintegrated over this time. She has a vivid memory of confronting him one day, when he was sitting at his dining room table. “I said to him, ‘Why do you treat me so different from Sandi? What is it? Am I not your child?’”

Joe looked out the window, Cheryl remembers. Then he looked down at his coffee cup and said, “Well, I wouldn’t go so far as to say that.”

§

DNA TESTS, IF DONE RIGOROUSLY, are far more definitive than tattered forms in old cedar chests, and far more emotionally potent. The genetic genealogy industry had barely gotten off the ground before scientists, sociologists, and ethicists were debating its societal impact—for better and for worse.

Early concerns focused on accuracy. All of the tests—whether they look at the Y chromosome, autosomal chromosomes, or other types of DNA—work in essentially the same way. They screen the billions of letters of a person’s DNA for a certain number of markers, and then compare that combination of markers with those found in reference samples taken from thousands of people living in various regions of the world. Test accuracy, then, starts with two things: the number of markers analyzed, and the size and selection of the comparison set of samples.

When genetic genealogy debuted, the technology cost many times what it does today. So the first tests screened a relatively small number of markers, leading to a crude measure of ancestry. The first test that Blaine Bettinger bought, from a now-defunct company called DNAPrint Genomics, screened his autosomal chromosomes for just 71 markers, and used those to estimate his ties to four broad ethnic groups: 88 percent “Indo-European,” 12 percent “East-Asian,” and zero percent “Native-American” and “African.”

“Those early autosomal tests were sort of wildly inaccurate,” Bettinger says. The subsequent tests he bought showed that his Honduran ancestors had both Native American and African roots.

Today’s genetic tests can probe many more markers, making them much more accurate. 23andMe uses around half a million markers, on all 23 chromosomes, to probe each customer’s ancestry. But many other companies continue to use only a small number of markers, and none make their reference databases or methodology transparent to customers. As one group of scholars wrote in a 2009 paper in Science: “Genetic ancestry tests fall into an unregulated no-man’s land, with little oversight and few industry guidelines to ensure the quality, validity, and interpretation of information sold.”

Even if all the tests were completely accurate, they’d still pose big philosophical questions: How much weight do individuals give to genes when forming ethnic, racial, and religious identities? How much weight should they give to DNA?

Cheryl, when pressed, acknowledges that she is some combination of her innate genetic predisposition, traumatic upbringing, and six decades of life experience. But like many people, she seems to give special weight to her genes.

As she posted to a site called Cousin Connect shortly after finding out that Joe wasn’t her father: “I want nothing from anyone [except] to know what blood line flows through my veins, my children, and grandchildren’s.”

Many people find religious and cultural identities in their DNA. Take Andrea, a 35-year-old adoptee. When she was less than two years old, her biological parents put her and two older brothers in foster care. The children were soon adopted, but their new parents were alcoholics, and they had messy and difficult childhoods.

At 16, Andrea left home and, a few years later, began searching for her birth parents.

She found her father’s profile on a dating website and called him. “That was a really hard phone call, because he was not interested in me,” she says. He also told her about her biological mother, and some searching revealed that she had died.

“It was very upsetting.”

But Andrea was profoundly uplifted by the results of her DNA test, which she bought from 23andMe about a year ago. The test indicated that she’s approximately one-quarter Ashkenazi Jewish. “That was like, the shock of all shocks,” she says. Though she is a practicing Christian, she has felt strong ties to Jewish culture since college, where she was a religious studies major. “I was very, very drawn to Jewish studies classes, I took biblical Hebrew, and always wanted to go to Israel,” she says.

Finding out that she had genetic roots to Judaism was bittersweet, she says, because she would have liked to have grown up in the Jewish culture. She’s making up for it now by reading all she can about Jewish history. “I will sit at home and watch documentaries on YouTube about Jerusalem,” she says, laughing. “I love it. And it’s so fascinating to me—the personal connection I had [with Judaism] even before I knew, and the one that continues now in my life.”

That kind of emotional connection, the “A-ha!” moment, is what Cheryl has been searching for all her life. She’s always wondered why she and her children don’t look like her sisters, Vivian, or Joe. Her son, Travis Whittle, has curly hair and a gregarious personality that Cheryl says resembles no one else in the family.

She wonders about her own traits and predispositions, too. She has had several bouts of depression over the years, and is almost always anxious. “I know that my mother had some depression. But I wonder if my father might have had some problems, too.”

It wasn’t enough to know that Joe wasn’t her father. To feel whole, she had to know who her real father was. “You know how in the Bible it says, so-and-so begot so-and-so begot so-and-so?” Cheryl says. “If you leave out a begot, there’s something missing. It doesn’t quite fit.”

§

CHERYL’S FERVENT HUNT FOR the mystery man responsible for half of her genetic identity has consumed much of her time over the past three years.

She didn’t have much luck browsing the Relative Finder section of the 23andMe website, which compared her DNA to that of the other people in the company’s database who had opted to share with the community. Her only genetic matches were estimated fourth, fifth, and sixth cousins—nowhere near close enough to trace back to her father.

In June of 2010, Cheryl bought Family Tree DNA’s genetic test, “to fish in more ponds,” as she puts it. The $293 cheek-swab test gave her access to all of the people in Family Tree DNA’s database, any one of which could have been a match. Unfortunately, though, she caught no fish.

Cheryl’s search went cold for nearly two years before picking up in April 2012, when CeCe Moore put her in touch with Diane Harman-Hoog of Redmond, Washington. Diane has spent her retirement years—“17 hours a day, 7 days a week,” she says—as a genealogy “search angel,” helping hundreds of people, mostly adoptees, figure out their family mysteries at no charge. Diane had just started to add genetic results into her search methods, and was eager to look into Cheryl’s case.

By pooling information from various sites, Diane created a spreadsheet showing more than five hundred people who shared some of Cheryl’s DNA. Each line of the spreadsheet gave the person’s surname and the precise chromosome location where their DNA matched Cheryl’s. But, big as it was, Diane’s spreadsheet didn’t identify any useful leads. “Diane wrote me and said, ‘Cheryl, we do not have enough. Your matches aren’t close enough yet,’” Cheryl recalls.

That was August 2012. Just a few months earlier, Ancestry.com—the largest genealogy company in the world—began selling its first autosomal DNA test. Cheryl bought one in September, for $99, to try her luck in yet another pond. Eight weeks later she had her matches: Nobody was closer than a fourth cousin.

During this lull in her search, Cheryl says, she had a profound spiritual experience. “One morning I got out of bed—and this sounds crazy probably, but, you know, I believe in God. I was just feeling real down about it. But then something inside of me said, ‘You will find your father.’ And so I was clinging to that. I knew that it’s just a matter of not losing hope.”

One February morning this year, Cheryl received a note via Ancestry.com’s internal message service from a woman named Jeannette Morrison, a genealogy hobbyist who had taken the test to expand her tree. She had identified Cheryl as a possible second cousin. Cheryl wrote back immediately, and updated Diane about the new lead. A second cousin match, Cheryl knew, could be a very big deal.

While Ancestry.com’s test will estimate the relatedness of two people, it doesn’t allow customers to compare their genetic data chromosome by chromosome. And Cheryl couldn’t tell from Jeannette’s family tree whether they were related through Cheryl’s mother or father. So Cheryl bought another 23andMe kit and sent it to Jeannette’s house in Ohio.

When Jeannette’s test results came back, in April, Cheryl discovered that Jeannette did not share any DNA with Sandi. In other words, Jeannette was exactly what Cheryl had been praying for: a solid lead to her biological father.

23andMe showed that Jeannette and Cheryl shared 17 segments of their DNA—including, crucially, two bits of the X chromosome. Through logical inferences and painstaking searches—comparing trees, geographical locations, birth and death dates—Diane found one of Jeannette’s relatives, Joseph Parker, who was about the right age and had lived in the right place to be Cheryl’s biological father.

Joseph had died in 1987, leaving behind a son, Joseph Jr., and one daughter, Effie Jane, who lived in Richmond. According to Diane’s analysis, Effie Jane could either be Cheryl’s first cousin once removed or her half-sister.

There was only one way to find out.

§

GENETIC GENEALOGY IS PART OF the much broader cultural trend of uploading personal data to the cloud. We willingly flaunt photos, videos, and demographic information on social media—Facebook, Twitter, Flickr, match.com—and give our credit card and social security numbers to banking and retail sites. Even seemingly private data—emails, cell phone records, internet browsing patterns—is actually, we’re learning, under government surveillance.

What’s fascinating about genetic genealogy is that it brings together two very different perspectives on privacy. DNA is arguably as personal as it gets. It’s an individual’s unique code of life. That’s why, among doctors and healthcare workers, genetic data is subject to strict privacy regulations.

The traditional genealogy community, on the other hand, is all about sharing—sharing family trees, sharing documents, sharing stories. “The only way you can connect with people is with some loss of privacy,” says Yaniv Erlich, a geneticist at the Whitehead Institute for Biomedical Research in Cambridge, Massachusetts.

Erlich is a world expert on genetic privacy. Earlier this year, his team caused a stir among medical researchers with a study in Science showing that supposedly anonymous participants in genetic research studies can be identified using simple software and an internet hook-up.

But Erlich is also an avid genealogist.

In the past few years he has bought DNA tests from 23andMe and Family Tree DNA, and has chosen to upload his genetic data to their databases. In doing so, he discovered that he carries the Cohen profile on his Y chromosome, confirming what had been passed down through his family’s oral tradition. He also found a fifth cousin whom he later met at a family reunion in Poland. That cousin grew up as a Christian, but because of his genetic discoveries is converting to Judaism. “It touches people, what they find in their DNA,” Erlich says. “I think it’s wonderful.”

The core privacy tension in genetic genealogy, Erlich notes, is that your DNA is not yours alone. “By putting your data out there,” Erlich says, “you’re not only sacrificing your own privacy but also the privacy of people who are connected to you, because you share DNA.”

In June, the Times, a British newspaper, ran a front-page story with the headline, “Revealed: The Indian Ancestry of William.” Two distant cousins of Prince William had their DNA tested with a company called BritainsDNA and discovered that they carried a rare set of markers that had previously been found in only 14 people: 13 Indian and one Nepalese. Because the DNA in question passes only from mother to child, and the cousins shared a great-great-great-great grandmother with William’s mother Diana, they could infer that the heir to the throne also has these Indian roots.

That particular bit of trivia is only important if you’re in the business of selling newspapers, as commentator Alex Hern pointed out in the New Statesman. But what if the genetic intel hadn’t been so silly? As Hern put it, “There is a wider issue at stake here, which is that the story reveals information about the genetic make-up of someone who has not consented to any DNA tests.”

The loose definition of genetic privacy, of course, is what allows people like Cheryl to solve their life mysteries. Cheryl’s cousin Jeannette, by agreeing to a DNA test, opened up the possibility of Cheryl identifying her real father and his descendants—regardless of whether any of them wanted segments of their DNA posted on a public database.

This risk—that relatives may be harmed in some way by the sharing of their DNA—has led some to argue that the decision to share is not an individual’s to make. In 2010, Henry Louis Gates, Jr. asked 12 celebrities to get DNA tests for his television show about genealogy, Faces of America. The novelist Louise Erdrich was the only one to refuse. Erdrich’s maternal grandfather was a chief of the Turtle Mountain Chippewas, a Native American tribe in North Dakota, and Erdrich is also an enrolled member. As Erdrich explained to Gates regarding the DNA test:

“It wouldn’t do me any harm, but when I asked my extended family about this—and I did go to everyone—I was told, ‘It’s not yours to give, Louise.’”

Legally, however, genetic testing is an individual decision. And unlike Erdrich, many people only consider privacy repercussions when they’re suddenly facing them.

MIKE TAFFE HADN’T GIVEN MUCH THOUGHT to violating his extended family’s privacy when he sent a tube of his spit to 23andMe in early 2012. Taffe is a neuroscientist at the Scripps Research Institute in La Jolla, California, and he was interested in the medical risk markers he might be carrying. As an adoptee, he would also become interested in the company’s Relative Finder service.

For Taffe’s first few months on the service, the closest matches were not close: third or fourth cousins.

Then a first cousin popped up: an African American man named Chris. Taffe wrote to Chris using 23andMe’s messaging service, explaining that he was looking for his father, who was, according to records from the adoption agency, a Puerto Rican man born in the 1940s. Chris said that didn’t ring any bells, and their correspondence ended soon after.

Taffe let it go. Then a year later, the company dropped its prices to $99 per kit, spurring him to send in spit samples of his three children.

Back on the site, he was reminded of his message to Chris and started snooping around online. He found Chris’s Facebook page, which was open to public viewing, and clicked through his photos. From these, he spotted Chris’s mother. (They had to be related through her, Taffe reasoned, because 23andMe had shown that they didn’t share Y chromosome DNA.) Then on the Facebook page of his presumed aunt, Taffe found some of her high school photos and her name.

After some more judicious Googling, he found an obituary for his aunt’s brother. No other siblings were mentioned. This man, Taffe thought, could very well be his biological father. The obit named the man’s three surviving children, and one of them, Cliff, had a Facebook page.

Taffe was immediately struck by Cliff’s photo: They shared a nose. “It was like, whoa, dude,” Taffe says. Looking through his presumed half-brother’s photos, Taffe saw a few that Cliff had posted of his father around the time of his death. The nose was the same.

This summer, Taffe took the plunge and sent a Facebook message to Cliff. It was vague, saying only that he was interested in ancestry and that he thought they might be related. Cliff never responded. If he ever does, Taffe isn’t sure how much more he would disclose.

“People have a right to privacy, even from their relatives,” Taffe says.

He has started a discussion about these issues on a 23andMe forum targeted at adoptees. “Obviously a lot of the adoptees have decided that their right to know trumps any other possible consideration. I do not, at present, agree with this.”

The 23andMe community forums, open to any 23andMe customer, are filled with gripping stories of people who are looking for their parents. They’re usually adoptees who bought the 23andMe test for this purpose. But there are also stories of people who inadvertently discovered that their parents weren’t who they said they were.

One such story comes from Terry, a college student whose world was rocked by the matches she found on 23andMe. At the top of the list was a man, whom she calls John Doe, who shares 28 percent of his DNA with her. 23andMe suggested that he was her grandfather, but that didn’t make sense given what she knew about her family.

Terry sent John Doe a message through 23andMe and learned that he is 23 years old, ruling out the grandfather theory. After hunting around online, she discovered that it’s not only her grandparents who share one-quarter of her DNA; John Doe could also be her uncle, half-brother, or double first cousin. Since she didn’t share any X chromosome pieces with John Doe, she figured he was a half-brother through her father.

Terry called her mother and asked her what she made of John Doe. As Terry explained on the forum, after initially seeming baffled her mother tried to deter her from pursuing John Doe. Over the next few days, dozens of other community members chimed in with advice and sympathy. One member asked Terry if she had considered the possibility that her parents had used a sperm donor. She had. Her parents were married for nine years before her birth, she wrote, so it was possible they had had fertility issues. After learning about John Doe, Terry had asked her mother directly if they had used a sperm donor — a suggestion her mother described as “insulting.” Terry believed her mother at first. But doubts crept in, partly because Terry’s ancestry composition did not match her father’s ethnicity.

John Doe, for his part, didn’t seem to understand genetics. He told Terry he thought they were cousins, and that it was just a “coincidence” that they shared so much DNA. Terry gently reminded him that first cousins share 12.5 percent of their DNA, on average — not 28 percent.

Eventually, Terry’s parents confessed: They had indeed used a sperm donor. Despite the new information, John Doe was in denial, continuing to say that he was definitely his father’s son.

Terry’s story illustrates the web of privacy concerns that these tests can create. Her results challenged the privacy of her parents, her half-brother, and her half-brother’s parents. Terry’s results may ultimately affect the privacy of her biological father as well. She has sent messages to several people the database identified as her first cousins, thinking that they might know an uncle who donated sperm.

Terry would have preferred that her parents had told her the truth, but she understands why they didn’t. In fact, she’s now keeping the same secret from a younger sibling who was conceived with a different sperm donor. As someone on the forum pointed out, Terry will have to keep her sibling away from 23andMe.

23ANDME SPELLS OUT SEVERAL PRIVACY ISSUES in its Terms of Service agreement. In the fifth of 28 sections, titled “Risks and Considerations Regarding 23andMe Services,” the company states several stark facts in bold type, including:

• Once you obtain your Genetic Information, the knowledge is irrevocable.
• You may learn information about yourself that you do not anticipate.
• Genetic Information you share with others could be used against your interests.

In a different document, called the Privacy Statement, the company highlights possible consequences of sharing genetic information. “Personal Information, once released or shared, can be difficult to contain,” it reads. “It is incumbent upon customers to share Personal Information only with people they know and trust.”

Family Tree DNA and Ancestry.com have similar agreements, though they’re less explicit about these intangible risks. It’s impossible to know, of course, how many customers ever read them. “We’d like to believe that everyone who’s accepted the terms has actually read them, but in practice we know that that’s not always the case,” says Catherine Afarian, a spokesperson for 23andMe.

As the company’s database grows, there are bound to be more people who find out family secrets. “It’s a reflection of what actually happens in our society,” Afarian says. Privacy is a top concern for the company, she adds, which is why the consent forms and privacy documents are written in clear and explicit language. “We try to be really upfront.” For now, these issues haven’t caused any legal problems or customer upsets, she says, and if they do, the company’s privacy policy “could certainly evolve over time.”

Nor is privacy the only area where DNA testing’s ethics are considered a work in progress. In November, 23andMe came under fire from the US Food and Drug Administration, for the part of its service that offers medical risk profiles. Customer results, delivered online with brief explanations of genetic indicators, such as Sandi’s discovery that she was at high risk of cancer, therefore fall under the FDA’s regulatory regime.

The agency accused the company of failing to cooperate fully over a period of four years, and of failing to communicate at all for more than six months. And although 23andMe rebuffed the claims, the company has now stopped marketing its services, and faces a class action lawsuit from disgruntled customers.

Cheryl has bought 16 kits from 23andMe, two for her (after her first one, she bought an updated version) and 14 on behalf of her friends and family. For each of them, she asked their permission and explained what she wanted to use the data for. But nobody ever read the forms.

Still, not everybody is interested in playing the genetic genealogy game. When Cheryl asked her daughter Wendy’s husband, Dennis Plear, if he’d like to test, he flatly refused, and also forbade Wendy from having their children tested. Dennis, a disabled Navy veteran who is half African American and half German, is wary of such personal information winding up in the hands of the government.

“When you start giving out samples of your DNA, you’re opening the door for other people to be in your business,” he says.

When told that these companies purport to be keeping the information secure and out of reach of third parties, Dennis wasn’t impressed. “If you have it on file, on the internet, the government has access to it one way or another.”

But for every customer deterred by privacy concerns, there seem to be many more who don’t care. Genetic genealogy databases are growing every day, and that pace will quicken as the costs of genetic sequencing drop.

It’s plausible that in the not-too-distant future, we’ll all be identifiable in genetic databases, whether through our personal contribution or that of our relatives. Some ethicists have called for federal legislation to restrict these kinds of collections. But many scientists, including Erlich, are against that idea, pointing out that genetic databases can be used for good—not only for genealogy, but for medical discoveries. “We cannot change the course of technology,” Erlich says.

Instead, he says, regulation should focus on preventing harm to individuals. For example, the Genetic Information Nondiscrimination Act (GINA), passed in the United States in 2008, says that health insurers and employers cannot use an individual’s genetic information to deny medical coverage or to make employment decisions. “GINA was a good step forward,” Erlich says. “It’s not about having the information— it’s about not abusing it.”

§

FROM HER TRAINING AS A NURSE, Cheryl is aware of the need to respect other people’s privacy. So in May of this year, when she first reached out to Effie Jane, she was wary of divulging too much, saying vaguely that she was interested in the Parker family line.

About a week later, Effie Jane wrote her back saying she would be happy to talk to her. Over the next couple of weeks, Cheryl and Effie Jane shared many emails and phone calls, and discovered that they had a lot in common.

“I never felt like I was a part of my family,” Effie Jane says. Her ruddy complexion and crystal-blue eyes look like Cheryl’s, but that’s the extent of their physical resemblance. Their childhoods, though, were similarly traumatic. Effie Jane started having seizures at six months old, and at 14 her mother died. Her father, Joseph Parker, like Cheryl’s adoptive father, Joe, was cold and indifferent, and “may have had a little mental imbalance,” Effie Jane says. After her mother died, his personality changed; for example, he instructed her to start calling him “Mr. Parker” rather than “Daddy.”

Effie Jane has felt rejected her whole life, but found acceptance in Cheryl. “I had always prayed for a sister,” she says. After meeting for the first time at Panera, the women kept talking, and slowly began to broach the sensitive subject on both of their minds: If they were indeed sisters, then how would Vivian have known Mr. Parker?

Mr. Parker was a train engineer, and would often be away from home for days at a time. Vivian frequently took the same train he worked on to vacation with her family. So she could have known him for years. Or perhaps she met him just once. Maybe they were friendly; or, who knows, maybe she was raped. These were the theories that Cheryl and Effie Jane tossed around. They wouldn’t know for sure until the test results came back. But Cheryl felt in her gut that this was the one.

“I’m feeling really comfortable with where I’m going with this. I feel like I’m going to find answers soon.” That’s what Cheryl said to me during our first phone interview, in early July. She had been anxiously tracking the shipment of Effie Jane’s kit and knew that it had arrived at 23andMe that Saturday. She had just a few weeks to wait.

Cheryl and Effie Jane’s relationship blossomed over that period. As Cheryl described in an email to Sandi, “There are a lot of things about Effie that seem familiar to me, as though we are of the same cloth.”

Cheryl and Effie Jane went to the cemetery where Mr. Parker is buried and snapped photos of his plaque. They also attended an all-day church workshop put on by Cheryl’s son, Travis, about how to deal with rejection. “My mom has a hard time believing that we accept her,” Travis says. “But that’s because of the deck of cards that she’s playing with.”

On July 23, Cheryl wrote to me with the surprising verdict: Effie Jane was not her sister. She was not even a first cousin once removed, as Diane had guessed. The genetic test estimated them as fourth cousins, at most, with just two shared segments of chromosome.

I asked Cheryl how she was feeling about it, and her response read, to me, like she was trying to hide her disappointment: “I am good! I am frustrated, as now we will have to find someone else to test, and wait yet again!”

Effie Jane took the news in stride. “I didn’t cry. I didn’t have that feeling of being deserted by her,” she told me. And the fact that you don’t share much DNA, does that change how you feel about your relationship, I asked. “As far as I’m concerned, she’s my little sister.”

Cheryl, too, says she will probably be lifelong friends with Effie Jane. But there was still a hole in her personal history. And it needed to be filled.

§

EFFIE JANE’S TEST RESULTS WEREN’T WHAT Cheryl was hoping for, but they did help her and Diane to focus on the relevant part of Jeannette’s family tree. Diane quickly identified another man as Cheryl’s potential father, and Cheryl tracked down the phone number of the man’s daughter, Rose.

Cheryl called Rose one morning in late July. The conversation went surprisingly well. “She said, ‘Oh, are you saying you might be my sister? Oh, I hope you might be my sister!’” Cheryl recalls. Rose agreed to take a 23andMe test, and they tentatively planned to get together the following day to do the spitting.

Late that evening, Cheryl had an upsetting phone call with Rose. The DNA test had been weighing on her the entire day, Rose said, and she had decided she didn’t want to tarnish the memory of her father by taking it. “Of course I was in tears,” Cheryl told me the next morning. “It’s like another rejection.”

Cheryl says she understands how Rose must be feeling. “I am nobody to them.” She is now trying to decide whether she should reach out to Rose’s late brother’s wife, to ask if her children might get tested. Cheryl knows that Rose would be upset if she did that; on the other hand, there may be no other way to find out who her father is. “It will make me feel more complete to know who I came from.”

Though she doesn’t fully realize it, Cheryl is playing out the hypothetical scenario painted by bioethicists: How does one person’s right to know stack up against another’s right to privacy?

WHILE SHE STRUGGLES WITH her search for a new family, Cheryl is trying to mend the tears that her genetic testing has created among the relatives she already has.

In June of last year, the Village News, the local newspaper of Chester, ran a story about Cheryl’s search for her biological father. The article included several factual inaccuracies, as well as a wildly inaccurate description of Joe as a man who “stepped up to the plate making Vivian, as they said then, an honest woman.”

All of Cheryl’s sisters were upset by the article. Sandi was upset because of what it implied about Vivian. “[It] shed a somewhat dim light of my mom and some of the report was inaccurate and fabricated,” she told me. “I only want my parents’ legacy to be respected.”

On the other hand, Cheryl’s older siblings, her half-sisters by Vivian’s first husband, were crushed by the description of Joe as an upstanding husband and father. “My sisters were furious. They really wanted to go for blood,” says one of them, JoAnn Lear, who, unlike Cheryl, is tall and boisterous, with a tell-it-like-it-is demeanor. They thought about suing the newspaper, but decided the damage had already been done.

I met JoAnn this past August, during a weekend I spent with Cheryl and Dickie at their home. On Saturday afternoon Cheryl organized a family reunion of sorts, so that I could meet her children, grandchildren, and Effie Jane. By the time I arrived, I had been talking to Cheryl on and off for several weeks about her hunt for her biological father. I was expecting to observe strained or awkward family relationships—holes that Cheryl was trying to fill with her genetic search. Instead I was struck by the love, humor, and openness of Cheryl’s family, new and old. When Effie Jane—Cheryl’s newest fourth cousin—arrived, she hugged Cheryl and gave her a gift: a sand dollar. It was a “Holy Ghost Shell,” she explained, with the star in the center representing the Star of Bethlehem.

Sandi wasn’t at the reunion—her colorectal cancer had by then spread to her lungs, and it was uncomfortable for her to leave her house or even to speak on the phone. (All of our discussions for this article happened through Facebook chats and emails; Sandi died on September 6, 2013.) Cheryl’s three older sisters weren’t there, either. None of them had talked to Cheryl much since the infamous Village News article had come out.

JoAnn, though, had agreed to meet with us the next day.

So Cheryl and I made the pretty, winding drive from her riverfront home to the suburbs of Richmond, where JoAnn lives, and picked her up for lunch and a ride around the small towns where they grew up. Cheryl and JoAnn had lost touch in recent years, but had recently reunited to care for Sandi.

After about an hour and a half of driving around, we pulled into a quiet road called Mason Avenue, in Chester. “It was one of these two,” said Cheryl from the back seat, pointing to two small houses. Each house used to hold two apartments, and one of them is where Cheryl, Sandi, Joe, and Vivian once lived.

This is the place, Cheryl said, where Joe once pulled a gun on her and Sandi and threatened to kill them.